User login
Artificial intelligence in the office: Part 2
In the year since generative artificial intelligence (AI) software first began to emerge for use, the staggering pace and breadth of development has condensed years of growth and change into months and weeks. Among the settings where these tools may find the greatest straight-line relevance is private medical practice.
. A multitude of generative AI products with potential medical applications are now available, with new ones appearing almost weekly. (As always, I have no financial interest in any product or service mentioned in this column.)
Last month, I discussed ChatGPT, the best-known AI algorithm, and some of its applications in clinical practice, such as generating website, video, and blog content. ChatGPT can also provide rapid and concise answers to general medical questions, like a search engine – but with more natural language processing and contextual understanding. Additionally, the algorithm can draft generic medical documents, including templates for after-visit summaries, postprocedure instructions, referrals, prior authorization appeal letters, and educational handouts.
Another useful feature of ChatGPT is its ability to provide accurate and conversational language translations, thus serving as an interpreter during clinic visits in situations where a human translator is not available. It also has potential uses in clinical research by finding resources, formulating hypotheses, drafting study protocols, and collecting large amounts of data in short periods of time. Other possibilities include survey administration, clinical trial recruitment, and automatic medication monitoring.
GPT-4, the latest version of ChatGPT, is reported to have greater problem-solving abilities and an even broader knowledge base. Among its claimed skills are the ability to find the latest literature in a given area, write a discharge summary for a patient following an uncomplicated surgery, and an image analysis feature to identify objects in photos. GPT-4 has been praised as having “the potential to help drive medical innovation, from aiding with patient discharge notes, summarizing recent clinical trials, providing information on ethical guidelines, and much more.”
Bard, an AI “chat bot” introduced by Google earlier this year, intends to leverage Google’s enormous database to compete with ChatGPT in providing answers to medical questions. Bard also hopes to play a pivotal role in expanding telemedicine and remote care via Google’s secure connections and access to patient records and medical history, and “facilitate seamless communication through appointment scheduling, messaging, and sharing medical images,” according to PackT, a website for IT professionals. The company claims that Bard’s integration of AI and machine learning capabilities will serve to elevate health care efficiency and patient outcomes, PackT says, and “the platform’s AI system quickly and accurately analyzes patient records, identifies patterns and trends, and aids medical professionals in developing effective treatment plans.”
Doximity has introduced an AI engine called DocsGPT, an encrypted, HIPAA-compliant writing assistant that, the company says, can draft any form of professional correspondence, including prior authorization letters, insurance appeals, patient support letters, and patient education materials. The service is available at no charge to all U.S. physicians and medical students through their Doximity accounts.
Microsoft has introduced several AI products. BioGPT is a language model specifically designed for health care. Compared with GPT models that are trained on more general text data, BioGPT is purported to have a deeper understanding of the language used in biomedical research and can generate more accurate and relevant outputs for biomedical tasks, such as drug discovery, disease classification, and clinical decision support. Fabric is another health care–specific data and analytics platform the company described in an announcement in May. It can combine data from sources such as electronic health records, images, lab systems, medical devices, and claims systems so hospitals and offices can standardize it and access it in the same place. Microsoft said the new tools will help eliminate the “time-consuming” process of searching through these sources one by one. Microsoft will also offer a new generative AI chatbot called the Azure Health Bot, which can pull information from a health organization’s own internal data as well as reputable external sources such as the Food and Drug Administration and the National Institutes of Health.
Several other AI products are available for clinicians. Tana served as an administrative aid and a clinical helper during the height of the COVID-19 pandemic, answering frequently asked questions, facilitating appointment management, and gathering preliminary medical information prior to teleconsultations. Dougall GPT is another AI chatbot tailored for health care professionals. It provides clinicians with AI-tuned answers to their queries, augmented by links to relevant, up-to-date, authoritative resources. It also assists in drafting patient instructions, consultation summaries, speeches, and professional correspondence. Wang has created Clinical Camel, an open-source health care–focused chatbot that assembles medical data with a combination of user-shared conversations and synthetic conversations derived from curated clinical articles. The Chinese company Baidu has rolled out Ernie as a potential rival to ChatGPT. You get the idea.
Of course, the inherent drawbacks of AI, such as producing false or biased information, perpetuating harmful stereotypes, and presenting information that has since been proven inaccurate or out-of-date, must always be kept in mind. All AI algorithms have been criticized for giving wrong answers, as their datasets are generally culled from information published in 2021 or earlier. Several of them have been shown to fabricate information – a phenomenon labeled “artificial hallucinations” in one article. “The scientific community must be vigilant in verifying the accuracy and reliability of the information provided by AI tools,” wrote the authors of that paper. “Researchers should use AI as an aid rather than a replacement for critical thinking and fact-checking.”
In the year since generative artificial intelligence (AI) software first began to emerge for use, the staggering pace and breadth of development has condensed years of growth and change into months and weeks. Among the settings where these tools may find the greatest straight-line relevance is private medical practice.
. A multitude of generative AI products with potential medical applications are now available, with new ones appearing almost weekly. (As always, I have no financial interest in any product or service mentioned in this column.)
Last month, I discussed ChatGPT, the best-known AI algorithm, and some of its applications in clinical practice, such as generating website, video, and blog content. ChatGPT can also provide rapid and concise answers to general medical questions, like a search engine – but with more natural language processing and contextual understanding. Additionally, the algorithm can draft generic medical documents, including templates for after-visit summaries, postprocedure instructions, referrals, prior authorization appeal letters, and educational handouts.
Another useful feature of ChatGPT is its ability to provide accurate and conversational language translations, thus serving as an interpreter during clinic visits in situations where a human translator is not available. It also has potential uses in clinical research by finding resources, formulating hypotheses, drafting study protocols, and collecting large amounts of data in short periods of time. Other possibilities include survey administration, clinical trial recruitment, and automatic medication monitoring.
GPT-4, the latest version of ChatGPT, is reported to have greater problem-solving abilities and an even broader knowledge base. Among its claimed skills are the ability to find the latest literature in a given area, write a discharge summary for a patient following an uncomplicated surgery, and an image analysis feature to identify objects in photos. GPT-4 has been praised as having “the potential to help drive medical innovation, from aiding with patient discharge notes, summarizing recent clinical trials, providing information on ethical guidelines, and much more.”
Bard, an AI “chat bot” introduced by Google earlier this year, intends to leverage Google’s enormous database to compete with ChatGPT in providing answers to medical questions. Bard also hopes to play a pivotal role in expanding telemedicine and remote care via Google’s secure connections and access to patient records and medical history, and “facilitate seamless communication through appointment scheduling, messaging, and sharing medical images,” according to PackT, a website for IT professionals. The company claims that Bard’s integration of AI and machine learning capabilities will serve to elevate health care efficiency and patient outcomes, PackT says, and “the platform’s AI system quickly and accurately analyzes patient records, identifies patterns and trends, and aids medical professionals in developing effective treatment plans.”
Doximity has introduced an AI engine called DocsGPT, an encrypted, HIPAA-compliant writing assistant that, the company says, can draft any form of professional correspondence, including prior authorization letters, insurance appeals, patient support letters, and patient education materials. The service is available at no charge to all U.S. physicians and medical students through their Doximity accounts.
Microsoft has introduced several AI products. BioGPT is a language model specifically designed for health care. Compared with GPT models that are trained on more general text data, BioGPT is purported to have a deeper understanding of the language used in biomedical research and can generate more accurate and relevant outputs for biomedical tasks, such as drug discovery, disease classification, and clinical decision support. Fabric is another health care–specific data and analytics platform the company described in an announcement in May. It can combine data from sources such as electronic health records, images, lab systems, medical devices, and claims systems so hospitals and offices can standardize it and access it in the same place. Microsoft said the new tools will help eliminate the “time-consuming” process of searching through these sources one by one. Microsoft will also offer a new generative AI chatbot called the Azure Health Bot, which can pull information from a health organization’s own internal data as well as reputable external sources such as the Food and Drug Administration and the National Institutes of Health.
Several other AI products are available for clinicians. Tana served as an administrative aid and a clinical helper during the height of the COVID-19 pandemic, answering frequently asked questions, facilitating appointment management, and gathering preliminary medical information prior to teleconsultations. Dougall GPT is another AI chatbot tailored for health care professionals. It provides clinicians with AI-tuned answers to their queries, augmented by links to relevant, up-to-date, authoritative resources. It also assists in drafting patient instructions, consultation summaries, speeches, and professional correspondence. Wang has created Clinical Camel, an open-source health care–focused chatbot that assembles medical data with a combination of user-shared conversations and synthetic conversations derived from curated clinical articles. The Chinese company Baidu has rolled out Ernie as a potential rival to ChatGPT. You get the idea.
Of course, the inherent drawbacks of AI, such as producing false or biased information, perpetuating harmful stereotypes, and presenting information that has since been proven inaccurate or out-of-date, must always be kept in mind. All AI algorithms have been criticized for giving wrong answers, as their datasets are generally culled from information published in 2021 or earlier. Several of them have been shown to fabricate information – a phenomenon labeled “artificial hallucinations” in one article. “The scientific community must be vigilant in verifying the accuracy and reliability of the information provided by AI tools,” wrote the authors of that paper. “Researchers should use AI as an aid rather than a replacement for critical thinking and fact-checking.”
In the year since generative artificial intelligence (AI) software first began to emerge for use, the staggering pace and breadth of development has condensed years of growth and change into months and weeks. Among the settings where these tools may find the greatest straight-line relevance is private medical practice.
. A multitude of generative AI products with potential medical applications are now available, with new ones appearing almost weekly. (As always, I have no financial interest in any product or service mentioned in this column.)
Last month, I discussed ChatGPT, the best-known AI algorithm, and some of its applications in clinical practice, such as generating website, video, and blog content. ChatGPT can also provide rapid and concise answers to general medical questions, like a search engine – but with more natural language processing and contextual understanding. Additionally, the algorithm can draft generic medical documents, including templates for after-visit summaries, postprocedure instructions, referrals, prior authorization appeal letters, and educational handouts.
Another useful feature of ChatGPT is its ability to provide accurate and conversational language translations, thus serving as an interpreter during clinic visits in situations where a human translator is not available. It also has potential uses in clinical research by finding resources, formulating hypotheses, drafting study protocols, and collecting large amounts of data in short periods of time. Other possibilities include survey administration, clinical trial recruitment, and automatic medication monitoring.
GPT-4, the latest version of ChatGPT, is reported to have greater problem-solving abilities and an even broader knowledge base. Among its claimed skills are the ability to find the latest literature in a given area, write a discharge summary for a patient following an uncomplicated surgery, and an image analysis feature to identify objects in photos. GPT-4 has been praised as having “the potential to help drive medical innovation, from aiding with patient discharge notes, summarizing recent clinical trials, providing information on ethical guidelines, and much more.”
Bard, an AI “chat bot” introduced by Google earlier this year, intends to leverage Google’s enormous database to compete with ChatGPT in providing answers to medical questions. Bard also hopes to play a pivotal role in expanding telemedicine and remote care via Google’s secure connections and access to patient records and medical history, and “facilitate seamless communication through appointment scheduling, messaging, and sharing medical images,” according to PackT, a website for IT professionals. The company claims that Bard’s integration of AI and machine learning capabilities will serve to elevate health care efficiency and patient outcomes, PackT says, and “the platform’s AI system quickly and accurately analyzes patient records, identifies patterns and trends, and aids medical professionals in developing effective treatment plans.”
Doximity has introduced an AI engine called DocsGPT, an encrypted, HIPAA-compliant writing assistant that, the company says, can draft any form of professional correspondence, including prior authorization letters, insurance appeals, patient support letters, and patient education materials. The service is available at no charge to all U.S. physicians and medical students through their Doximity accounts.
Microsoft has introduced several AI products. BioGPT is a language model specifically designed for health care. Compared with GPT models that are trained on more general text data, BioGPT is purported to have a deeper understanding of the language used in biomedical research and can generate more accurate and relevant outputs for biomedical tasks, such as drug discovery, disease classification, and clinical decision support. Fabric is another health care–specific data and analytics platform the company described in an announcement in May. It can combine data from sources such as electronic health records, images, lab systems, medical devices, and claims systems so hospitals and offices can standardize it and access it in the same place. Microsoft said the new tools will help eliminate the “time-consuming” process of searching through these sources one by one. Microsoft will also offer a new generative AI chatbot called the Azure Health Bot, which can pull information from a health organization’s own internal data as well as reputable external sources such as the Food and Drug Administration and the National Institutes of Health.
Several other AI products are available for clinicians. Tana served as an administrative aid and a clinical helper during the height of the COVID-19 pandemic, answering frequently asked questions, facilitating appointment management, and gathering preliminary medical information prior to teleconsultations. Dougall GPT is another AI chatbot tailored for health care professionals. It provides clinicians with AI-tuned answers to their queries, augmented by links to relevant, up-to-date, authoritative resources. It also assists in drafting patient instructions, consultation summaries, speeches, and professional correspondence. Wang has created Clinical Camel, an open-source health care–focused chatbot that assembles medical data with a combination of user-shared conversations and synthetic conversations derived from curated clinical articles. The Chinese company Baidu has rolled out Ernie as a potential rival to ChatGPT. You get the idea.
Of course, the inherent drawbacks of AI, such as producing false or biased information, perpetuating harmful stereotypes, and presenting information that has since been proven inaccurate or out-of-date, must always be kept in mind. All AI algorithms have been criticized for giving wrong answers, as their datasets are generally culled from information published in 2021 or earlier. Several of them have been shown to fabricate information – a phenomenon labeled “artificial hallucinations” in one article. “The scientific community must be vigilant in verifying the accuracy and reliability of the information provided by AI tools,” wrote the authors of that paper. “Researchers should use AI as an aid rather than a replacement for critical thinking and fact-checking.”
Neoadjuvant advantages: Treating locally advanced lung cancer
Many of you saw the press release from Merck announcing that their randomized trial comparing chemo with chemo plus pembrolizumab in the neoadjuvant setting led to improved event-free survival and also improved pathologic complete response rate.
This comes in addition to the data from the AstraZeneca trial with durvalumab saying they’ve already achieved their endpoint of higher pathologic complete response rate vs. chemotherapy alone and also the data with nivolumab from Bristol-Myers Squibb saying that nivolumab plus chemotherapy leads to a better event-free survival and a better pathologic complete response rate. That information has led to Food and Drug Administration approval for their regimen.
We’re running the table with these very positive data, and I think it’s just a sign that the approach is safe and effective.
A huge question has come up. I just came from a meeting of lung cancer experts asking what to do if you have a patient with a small tumor, for example, a 3-cm tumor. Do you recommend immediate surgery followed by adjuvant therapy, chemotherapy, and then a checkpoint inhibitor if appropriate? Or do you proceed with neoadjuvant therapy if appropriate? The truth is that it’s a very difficult decision.
We have overwhelming data that the neoadjuvant approach works for that patient. Please remember that this is a clinically staged patient. This is not the patient after their surgery, where I think we have a very clear path. We have adjuvant data and adjuvant trials for those patients.
For the patient who’s in your office with a small tumor or a small tumor and only hilar lymphadenopathy, the decision there isn’t data driven, but rather it is experience driven. The data that are out there right now suggest that neoadjuvant therapy is a better way to go. Why is that?
Well, I think that the first reason is that it is probably a better regimen. I think many of you saw the recent clinical trial by Patel and colleagues in the New England Journal of Medicine with melanoma. It was an interesting trial. They gave a checkpoint inhibitor for 18 doses after surgery for melanoma versus three doses of checkpoint inhibitor, surgery, and then 15 doses of the checkpoint inhibitor.
It was 18 doses versus 18 doses, with the only difference being the three doses before surgery. Lo and behold, the three doses before surgery led to a better event-free survival.
There are preclinical data in lung cancer demonstrating that the same thing is true. Tina Cascone published on that years ago. We could talk about why, but it appears that neoadjuvant is just better.
There are other advantages to it as well. I think a big one is that all the information shows that it’s better tolerated, so you’re more likely to give all the drug. You can see if the drug isn’t working, and you can stop the drug. Also, if the drug is causing a side effect, you can see whether it’s working or not and use that decision to stop. It’s different than when you’re giving a drug in the adjuvant setting where you don’t really know whether it is working or not.
I think that it’s time to change some of our standards. When patients appear with lung cancers other than tiny ones that might be detected through screening, you need to convene your multidisciplinary group. You need to weigh the pros and cons I think that it’s time to change some of our standards. When patients appear with lung cancers other than tiny ones that might be detected through screening, you need to convene your multidisciplinary group coming in. It’s already an FDA-approved regimen with nivolumab and chemotherapy, and I think we’re moving to making that our standard of care now.
The way to handle it today, though, is to convene your multidisciplinary panel about every patient other than those with the tiniest of lung cancers and put your heads together to see what the best treatment is for that patient.
Dr. Kris is professor of medicine, Weill Cornell Medicine, and the William and Joy Ruane Chair in Thoracic Oncology, Memorial Sloan Kettering Cancer Center, both in New York. He disclosed ties with Ariad Pharmaceuticals, AstraZeneca, Pfizer, PUMA, and Roche/Genentech.
A version of this article appeared on Medscape.com.
Many of you saw the press release from Merck announcing that their randomized trial comparing chemo with chemo plus pembrolizumab in the neoadjuvant setting led to improved event-free survival and also improved pathologic complete response rate.
This comes in addition to the data from the AstraZeneca trial with durvalumab saying they’ve already achieved their endpoint of higher pathologic complete response rate vs. chemotherapy alone and also the data with nivolumab from Bristol-Myers Squibb saying that nivolumab plus chemotherapy leads to a better event-free survival and a better pathologic complete response rate. That information has led to Food and Drug Administration approval for their regimen.
We’re running the table with these very positive data, and I think it’s just a sign that the approach is safe and effective.
A huge question has come up. I just came from a meeting of lung cancer experts asking what to do if you have a patient with a small tumor, for example, a 3-cm tumor. Do you recommend immediate surgery followed by adjuvant therapy, chemotherapy, and then a checkpoint inhibitor if appropriate? Or do you proceed with neoadjuvant therapy if appropriate? The truth is that it’s a very difficult decision.
We have overwhelming data that the neoadjuvant approach works for that patient. Please remember that this is a clinically staged patient. This is not the patient after their surgery, where I think we have a very clear path. We have adjuvant data and adjuvant trials for those patients.
For the patient who’s in your office with a small tumor or a small tumor and only hilar lymphadenopathy, the decision there isn’t data driven, but rather it is experience driven. The data that are out there right now suggest that neoadjuvant therapy is a better way to go. Why is that?
Well, I think that the first reason is that it is probably a better regimen. I think many of you saw the recent clinical trial by Patel and colleagues in the New England Journal of Medicine with melanoma. It was an interesting trial. They gave a checkpoint inhibitor for 18 doses after surgery for melanoma versus three doses of checkpoint inhibitor, surgery, and then 15 doses of the checkpoint inhibitor.
It was 18 doses versus 18 doses, with the only difference being the three doses before surgery. Lo and behold, the three doses before surgery led to a better event-free survival.
There are preclinical data in lung cancer demonstrating that the same thing is true. Tina Cascone published on that years ago. We could talk about why, but it appears that neoadjuvant is just better.
There are other advantages to it as well. I think a big one is that all the information shows that it’s better tolerated, so you’re more likely to give all the drug. You can see if the drug isn’t working, and you can stop the drug. Also, if the drug is causing a side effect, you can see whether it’s working or not and use that decision to stop. It’s different than when you’re giving a drug in the adjuvant setting where you don’t really know whether it is working or not.
I think that it’s time to change some of our standards. When patients appear with lung cancers other than tiny ones that might be detected through screening, you need to convene your multidisciplinary group. You need to weigh the pros and cons I think that it’s time to change some of our standards. When patients appear with lung cancers other than tiny ones that might be detected through screening, you need to convene your multidisciplinary group coming in. It’s already an FDA-approved regimen with nivolumab and chemotherapy, and I think we’re moving to making that our standard of care now.
The way to handle it today, though, is to convene your multidisciplinary panel about every patient other than those with the tiniest of lung cancers and put your heads together to see what the best treatment is for that patient.
Dr. Kris is professor of medicine, Weill Cornell Medicine, and the William and Joy Ruane Chair in Thoracic Oncology, Memorial Sloan Kettering Cancer Center, both in New York. He disclosed ties with Ariad Pharmaceuticals, AstraZeneca, Pfizer, PUMA, and Roche/Genentech.
A version of this article appeared on Medscape.com.
Many of you saw the press release from Merck announcing that their randomized trial comparing chemo with chemo plus pembrolizumab in the neoadjuvant setting led to improved event-free survival and also improved pathologic complete response rate.
This comes in addition to the data from the AstraZeneca trial with durvalumab saying they’ve already achieved their endpoint of higher pathologic complete response rate vs. chemotherapy alone and also the data with nivolumab from Bristol-Myers Squibb saying that nivolumab plus chemotherapy leads to a better event-free survival and a better pathologic complete response rate. That information has led to Food and Drug Administration approval for their regimen.
We’re running the table with these very positive data, and I think it’s just a sign that the approach is safe and effective.
A huge question has come up. I just came from a meeting of lung cancer experts asking what to do if you have a patient with a small tumor, for example, a 3-cm tumor. Do you recommend immediate surgery followed by adjuvant therapy, chemotherapy, and then a checkpoint inhibitor if appropriate? Or do you proceed with neoadjuvant therapy if appropriate? The truth is that it’s a very difficult decision.
We have overwhelming data that the neoadjuvant approach works for that patient. Please remember that this is a clinically staged patient. This is not the patient after their surgery, where I think we have a very clear path. We have adjuvant data and adjuvant trials for those patients.
For the patient who’s in your office with a small tumor or a small tumor and only hilar lymphadenopathy, the decision there isn’t data driven, but rather it is experience driven. The data that are out there right now suggest that neoadjuvant therapy is a better way to go. Why is that?
Well, I think that the first reason is that it is probably a better regimen. I think many of you saw the recent clinical trial by Patel and colleagues in the New England Journal of Medicine with melanoma. It was an interesting trial. They gave a checkpoint inhibitor for 18 doses after surgery for melanoma versus three doses of checkpoint inhibitor, surgery, and then 15 doses of the checkpoint inhibitor.
It was 18 doses versus 18 doses, with the only difference being the three doses before surgery. Lo and behold, the three doses before surgery led to a better event-free survival.
There are preclinical data in lung cancer demonstrating that the same thing is true. Tina Cascone published on that years ago. We could talk about why, but it appears that neoadjuvant is just better.
There are other advantages to it as well. I think a big one is that all the information shows that it’s better tolerated, so you’re more likely to give all the drug. You can see if the drug isn’t working, and you can stop the drug. Also, if the drug is causing a side effect, you can see whether it’s working or not and use that decision to stop. It’s different than when you’re giving a drug in the adjuvant setting where you don’t really know whether it is working or not.
I think that it’s time to change some of our standards. When patients appear with lung cancers other than tiny ones that might be detected through screening, you need to convene your multidisciplinary group. You need to weigh the pros and cons I think that it’s time to change some of our standards. When patients appear with lung cancers other than tiny ones that might be detected through screening, you need to convene your multidisciplinary group coming in. It’s already an FDA-approved regimen with nivolumab and chemotherapy, and I think we’re moving to making that our standard of care now.
The way to handle it today, though, is to convene your multidisciplinary panel about every patient other than those with the tiniest of lung cancers and put your heads together to see what the best treatment is for that patient.
Dr. Kris is professor of medicine, Weill Cornell Medicine, and the William and Joy Ruane Chair in Thoracic Oncology, Memorial Sloan Kettering Cancer Center, both in New York. He disclosed ties with Ariad Pharmaceuticals, AstraZeneca, Pfizer, PUMA, and Roche/Genentech.
A version of this article appeared on Medscape.com.
New kids on the block for migraine treatment and prophylaxis
This transcript has been edited for clarity.
Dear colleagues, I’m Hans-Christoph Diener from the Faculty of Medicine at the University of Duisburg-Essen in Germany.
CGRP receptor agonists
Let me start with the treatment of acute migraine attacks. Until recently, we had analgesics, nonsteroidal anti-inflammatory drugs like ibuprofen, ergot alkaloids, and triptans. There are new developments, which are small molecules that are antagonists at the calcitonin gene-related peptide (CGRP) receptor. At the moment, we have three of them: rimegepant 75 mg, ubrogepant 50 mg or 100 mg, and zavegepant (a nasal spray) 10 mg.
These are all effective and superior to placebo. The 2-hour pain-free rate is somewhere between 25% and 30%. They have very few side effects; these include a little bit of nausea, somnolence, nasopharyngitis, and for zavegepant, the nasal spray, taste disturbance. In indirect comparisons, the so-called gepants are about as effective as ibuprofen and aspirin, and they seem to be less effective than sumatriptan 100 mg.
Unfortunately, until now, we have no direct comparison with triptans and we have no data demonstrating whether they are effective in people where triptans do not work. The major shortcoming is the cost in the United States. The cost per tablet or nasal spray is somewhere between $80 and $200. This means we definitely need more studies for these gepants.
Migraine prophylaxis
Let me move to the prophylaxis of migraine with drugs. Previously and still, we have all medications like beta-blockers, flunarizine, topiramate, valproic acid, amitriptyline, and candesartan, and for chronic migraine, onabotulinumtoxinA. We have now 5 years’ experience with the monoclonal antibodies against CGRP or the CGRP receptor like eptinezumab, erenumab, fremanezumab, and galcanezumab.
These are all equally effective. They reduce migraine-days between 3 and 7 per month. They are effective both in episodic and chronic migraine, and most importantly, they are effective in people with medication overuse headaches. The 50% responder rates are somewhere between 40% and 60%, and there are no significant differences between the four monoclonal antibodies.
The major advantage is a very good tolerability profile; very few patients terminate treatment because of adverse events. There has been, with one exception, no direct comparison of the monoclonal antibodies with traditional migraine preventive drugs or onabotulinumtoxinA. The only exception is a trial that compared topiramate and erenumab, showing that erenumab was definitely more effective and better tolerated.
At the moment, the recommendation is to use these monoclonal antibodies for 12 months in episodic migraine and 24 months in chronic migraine and then pause. It usually turns out that between 50% and 70% of these patients need to continue the treatment. If they are not working, there is a possibility to switch between the monoclonal antibodies, and the success rate after this is somewhere between 15% and 30%.
Gepants were also developed for the prevention of migraine. Here, we have rimegepant 75 mg every other day or atogepant 60 mg daily. They are effective, but in indirect comparisons, they are less effective than the monoclonal antibodies. At present, we have no comparative trials with monoclonal antibodies or the traditional migraine preventive drugs.
Potential patients are those who have needle phobia or patients who do not respond to monoclonal antibodies. Again, the biggest shortcoming is cost in the United States. The cost per year for migraine prevention or prophylaxis is between $12,000 and $20,000.
Finally, we also had very exciting news. There is a new therapeutic approach via PACAP. PACAP is pituitary adenylate cyclase-activating polypeptide, which has similar biological actions as CGRP but with additional actions. It could very well be that people who do not respond to a monoclonal antibody would respond to a monoclonal antibody against PACAP.
At the congress, the first randomized, placebo-controlled trial with a monoclonal antibody against PACAP was presented. This monoclonal antibody was effective in a population of people in whom prior preventive therapy had failed. A phase 3 study is planned, and most probably the PACAP monoclonal could work in people who do not respond to monoclonal antibodies against CGRP.
Dear colleagues, we have now many choices for the acute treatment of migraine and migraine prophylaxis. We have new kids on the block, and we have to learn more about how to use these drugs, their benefits, and their shortcomings.
He has disclosed the following relevant financial relationships:Received honoraria for participation in clinical trials, contribution to advisory boards or oral presentations from: Abbott; Addex Pharma; Alder; Allergan; Almirall; Amgen; Autonomic Technology; AstraZeneca; Bayer Vital; Berlin Chemie; Bristol-Myers Squibb; Boehringer Ingelheim; Chordate; CoAxia; Corimmun; Covidien; Coherex; CoLucid; Daiichi-Sankyo; D-Pharm; Electrocore; Fresenius; GlaxoSmithKline; Grunenthal; Janssen-Cilag; Labrys Biologics Lilly; La Roche; 3M Medica; MSD; Medtronic; Menarini; MindFrame; Minster; Neuroscore; Neurobiological Technologies; Novartis; Novo Nordisk; Johnson & Johnson; Knoll; Paion; Parke-Davis; Pierre Fabre; Pfizer; Schaper and Brummer; Sanofi-Aventis; Schering-Plough; Servier; Solvay; Syngis; St. Jude; Talecris; Thrombogenics; WebMD Global; Weber and Weber; Wyeth; and Yamanouchi.
Dr. Diener is professor, department of neurology, Stroke Center-Headache Center, University Duisburg-Essen (Germany).
A version of this article appeared on Medscape.com.
This transcript has been edited for clarity.
Dear colleagues, I’m Hans-Christoph Diener from the Faculty of Medicine at the University of Duisburg-Essen in Germany.
CGRP receptor agonists
Let me start with the treatment of acute migraine attacks. Until recently, we had analgesics, nonsteroidal anti-inflammatory drugs like ibuprofen, ergot alkaloids, and triptans. There are new developments, which are small molecules that are antagonists at the calcitonin gene-related peptide (CGRP) receptor. At the moment, we have three of them: rimegepant 75 mg, ubrogepant 50 mg or 100 mg, and zavegepant (a nasal spray) 10 mg.
These are all effective and superior to placebo. The 2-hour pain-free rate is somewhere between 25% and 30%. They have very few side effects; these include a little bit of nausea, somnolence, nasopharyngitis, and for zavegepant, the nasal spray, taste disturbance. In indirect comparisons, the so-called gepants are about as effective as ibuprofen and aspirin, and they seem to be less effective than sumatriptan 100 mg.
Unfortunately, until now, we have no direct comparison with triptans and we have no data demonstrating whether they are effective in people where triptans do not work. The major shortcoming is the cost in the United States. The cost per tablet or nasal spray is somewhere between $80 and $200. This means we definitely need more studies for these gepants.
Migraine prophylaxis
Let me move to the prophylaxis of migraine with drugs. Previously and still, we have all medications like beta-blockers, flunarizine, topiramate, valproic acid, amitriptyline, and candesartan, and for chronic migraine, onabotulinumtoxinA. We have now 5 years’ experience with the monoclonal antibodies against CGRP or the CGRP receptor like eptinezumab, erenumab, fremanezumab, and galcanezumab.
These are all equally effective. They reduce migraine-days between 3 and 7 per month. They are effective both in episodic and chronic migraine, and most importantly, they are effective in people with medication overuse headaches. The 50% responder rates are somewhere between 40% and 60%, and there are no significant differences between the four monoclonal antibodies.
The major advantage is a very good tolerability profile; very few patients terminate treatment because of adverse events. There has been, with one exception, no direct comparison of the monoclonal antibodies with traditional migraine preventive drugs or onabotulinumtoxinA. The only exception is a trial that compared topiramate and erenumab, showing that erenumab was definitely more effective and better tolerated.
At the moment, the recommendation is to use these monoclonal antibodies for 12 months in episodic migraine and 24 months in chronic migraine and then pause. It usually turns out that between 50% and 70% of these patients need to continue the treatment. If they are not working, there is a possibility to switch between the monoclonal antibodies, and the success rate after this is somewhere between 15% and 30%.
Gepants were also developed for the prevention of migraine. Here, we have rimegepant 75 mg every other day or atogepant 60 mg daily. They are effective, but in indirect comparisons, they are less effective than the monoclonal antibodies. At present, we have no comparative trials with monoclonal antibodies or the traditional migraine preventive drugs.
Potential patients are those who have needle phobia or patients who do not respond to monoclonal antibodies. Again, the biggest shortcoming is cost in the United States. The cost per year for migraine prevention or prophylaxis is between $12,000 and $20,000.
Finally, we also had very exciting news. There is a new therapeutic approach via PACAP. PACAP is pituitary adenylate cyclase-activating polypeptide, which has similar biological actions as CGRP but with additional actions. It could very well be that people who do not respond to a monoclonal antibody would respond to a monoclonal antibody against PACAP.
At the congress, the first randomized, placebo-controlled trial with a monoclonal antibody against PACAP was presented. This monoclonal antibody was effective in a population of people in whom prior preventive therapy had failed. A phase 3 study is planned, and most probably the PACAP monoclonal could work in people who do not respond to monoclonal antibodies against CGRP.
Dear colleagues, we have now many choices for the acute treatment of migraine and migraine prophylaxis. We have new kids on the block, and we have to learn more about how to use these drugs, their benefits, and their shortcomings.
He has disclosed the following relevant financial relationships:Received honoraria for participation in clinical trials, contribution to advisory boards or oral presentations from: Abbott; Addex Pharma; Alder; Allergan; Almirall; Amgen; Autonomic Technology; AstraZeneca; Bayer Vital; Berlin Chemie; Bristol-Myers Squibb; Boehringer Ingelheim; Chordate; CoAxia; Corimmun; Covidien; Coherex; CoLucid; Daiichi-Sankyo; D-Pharm; Electrocore; Fresenius; GlaxoSmithKline; Grunenthal; Janssen-Cilag; Labrys Biologics Lilly; La Roche; 3M Medica; MSD; Medtronic; Menarini; MindFrame; Minster; Neuroscore; Neurobiological Technologies; Novartis; Novo Nordisk; Johnson & Johnson; Knoll; Paion; Parke-Davis; Pierre Fabre; Pfizer; Schaper and Brummer; Sanofi-Aventis; Schering-Plough; Servier; Solvay; Syngis; St. Jude; Talecris; Thrombogenics; WebMD Global; Weber and Weber; Wyeth; and Yamanouchi.
Dr. Diener is professor, department of neurology, Stroke Center-Headache Center, University Duisburg-Essen (Germany).
A version of this article appeared on Medscape.com.
This transcript has been edited for clarity.
Dear colleagues, I’m Hans-Christoph Diener from the Faculty of Medicine at the University of Duisburg-Essen in Germany.
CGRP receptor agonists
Let me start with the treatment of acute migraine attacks. Until recently, we had analgesics, nonsteroidal anti-inflammatory drugs like ibuprofen, ergot alkaloids, and triptans. There are new developments, which are small molecules that are antagonists at the calcitonin gene-related peptide (CGRP) receptor. At the moment, we have three of them: rimegepant 75 mg, ubrogepant 50 mg or 100 mg, and zavegepant (a nasal spray) 10 mg.
These are all effective and superior to placebo. The 2-hour pain-free rate is somewhere between 25% and 30%. They have very few side effects; these include a little bit of nausea, somnolence, nasopharyngitis, and for zavegepant, the nasal spray, taste disturbance. In indirect comparisons, the so-called gepants are about as effective as ibuprofen and aspirin, and they seem to be less effective than sumatriptan 100 mg.
Unfortunately, until now, we have no direct comparison with triptans and we have no data demonstrating whether they are effective in people where triptans do not work. The major shortcoming is the cost in the United States. The cost per tablet or nasal spray is somewhere between $80 and $200. This means we definitely need more studies for these gepants.
Migraine prophylaxis
Let me move to the prophylaxis of migraine with drugs. Previously and still, we have all medications like beta-blockers, flunarizine, topiramate, valproic acid, amitriptyline, and candesartan, and for chronic migraine, onabotulinumtoxinA. We have now 5 years’ experience with the monoclonal antibodies against CGRP or the CGRP receptor like eptinezumab, erenumab, fremanezumab, and galcanezumab.
These are all equally effective. They reduce migraine-days between 3 and 7 per month. They are effective both in episodic and chronic migraine, and most importantly, they are effective in people with medication overuse headaches. The 50% responder rates are somewhere between 40% and 60%, and there are no significant differences between the four monoclonal antibodies.
The major advantage is a very good tolerability profile; very few patients terminate treatment because of adverse events. There has been, with one exception, no direct comparison of the monoclonal antibodies with traditional migraine preventive drugs or onabotulinumtoxinA. The only exception is a trial that compared topiramate and erenumab, showing that erenumab was definitely more effective and better tolerated.
At the moment, the recommendation is to use these monoclonal antibodies for 12 months in episodic migraine and 24 months in chronic migraine and then pause. It usually turns out that between 50% and 70% of these patients need to continue the treatment. If they are not working, there is a possibility to switch between the monoclonal antibodies, and the success rate after this is somewhere between 15% and 30%.
Gepants were also developed for the prevention of migraine. Here, we have rimegepant 75 mg every other day or atogepant 60 mg daily. They are effective, but in indirect comparisons, they are less effective than the monoclonal antibodies. At present, we have no comparative trials with monoclonal antibodies or the traditional migraine preventive drugs.
Potential patients are those who have needle phobia or patients who do not respond to monoclonal antibodies. Again, the biggest shortcoming is cost in the United States. The cost per year for migraine prevention or prophylaxis is between $12,000 and $20,000.
Finally, we also had very exciting news. There is a new therapeutic approach via PACAP. PACAP is pituitary adenylate cyclase-activating polypeptide, which has similar biological actions as CGRP but with additional actions. It could very well be that people who do not respond to a monoclonal antibody would respond to a monoclonal antibody against PACAP.
At the congress, the first randomized, placebo-controlled trial with a monoclonal antibody against PACAP was presented. This monoclonal antibody was effective in a population of people in whom prior preventive therapy had failed. A phase 3 study is planned, and most probably the PACAP monoclonal could work in people who do not respond to monoclonal antibodies against CGRP.
Dear colleagues, we have now many choices for the acute treatment of migraine and migraine prophylaxis. We have new kids on the block, and we have to learn more about how to use these drugs, their benefits, and their shortcomings.
He has disclosed the following relevant financial relationships:Received honoraria for participation in clinical trials, contribution to advisory boards or oral presentations from: Abbott; Addex Pharma; Alder; Allergan; Almirall; Amgen; Autonomic Technology; AstraZeneca; Bayer Vital; Berlin Chemie; Bristol-Myers Squibb; Boehringer Ingelheim; Chordate; CoAxia; Corimmun; Covidien; Coherex; CoLucid; Daiichi-Sankyo; D-Pharm; Electrocore; Fresenius; GlaxoSmithKline; Grunenthal; Janssen-Cilag; Labrys Biologics Lilly; La Roche; 3M Medica; MSD; Medtronic; Menarini; MindFrame; Minster; Neuroscore; Neurobiological Technologies; Novartis; Novo Nordisk; Johnson & Johnson; Knoll; Paion; Parke-Davis; Pierre Fabre; Pfizer; Schaper and Brummer; Sanofi-Aventis; Schering-Plough; Servier; Solvay; Syngis; St. Jude; Talecris; Thrombogenics; WebMD Global; Weber and Weber; Wyeth; and Yamanouchi.
Dr. Diener is professor, department of neurology, Stroke Center-Headache Center, University Duisburg-Essen (Germany).
A version of this article appeared on Medscape.com.
Where do you stand on the Middle East conflict?
“What do you think about the whole Israel thing?”
That question came at the end of an otherwise routine appointment.
Maybe she was just chatting. Maybe she wanted something deeper. I have no idea. I just said, “I don’t discuss those things with patients.”
My answer surprised her, but she didn’t push it. She paid her copay, scheduled a follow-up for 3 months, and left.
As I’ve written before, I try to avoid all news except the local weather. The sad reality is that most of it is bad and there’s nothing I can really do about it. It only upsets me, which isn’t good for my mental health and blood pressure, and if I can’t change it, what’s the point of knowing? It falls under the serenity prayer.
Of course, some news stories are too big not to hear something. I pass TVs in the doctors lounge or coffee house, hear others talking as I stand in line for the elevator, or see blurbs go by when checking the weather. It’s not entirely unavoidable.
I’m not trivializing the Middle East. But, to me, it’s not part of the doctor-patient relationship any more than my political views are. You run the risk of driving a wedge between you and the person you’re caring for. If you don’t like their opinion, you may find yourself less interested in them and their care. If they don’t like your opinion on news, they may start to question your ability as a doctor.
That’s not what we strive for, but it can be human nature. For better or worse we often reduce things to “us against them,” and learning someone is on the opposite side may, even subconsciously, alter how you treat them.
That’s not good, so to me it’s best not to know.
Some may think I’m being petty, or aloof, to be unwilling to discuss nonmedical issues of significance, but I don’t see it that way. Time is limited at the appointment and is best spent on medical care. Something unrelated to the visit that may alter my objective opinion of a patient – or theirs of me as a doctor – is best left out of it.
I’m here to be your doctor, and to do the best I can for you. I’m not here to be a debate partner. Whenever a patient asks me a question on politics or news I always think of the Monty Python skit “Argument Clinic.” That’s not why you’re here. There are plenty places to discuss such things. My office isn’t one of them.
Dr. Block has a solo neurology practice in Scottsdale, Ariz.
“What do you think about the whole Israel thing?”
That question came at the end of an otherwise routine appointment.
Maybe she was just chatting. Maybe she wanted something deeper. I have no idea. I just said, “I don’t discuss those things with patients.”
My answer surprised her, but she didn’t push it. She paid her copay, scheduled a follow-up for 3 months, and left.
As I’ve written before, I try to avoid all news except the local weather. The sad reality is that most of it is bad and there’s nothing I can really do about it. It only upsets me, which isn’t good for my mental health and blood pressure, and if I can’t change it, what’s the point of knowing? It falls under the serenity prayer.
Of course, some news stories are too big not to hear something. I pass TVs in the doctors lounge or coffee house, hear others talking as I stand in line for the elevator, or see blurbs go by when checking the weather. It’s not entirely unavoidable.
I’m not trivializing the Middle East. But, to me, it’s not part of the doctor-patient relationship any more than my political views are. You run the risk of driving a wedge between you and the person you’re caring for. If you don’t like their opinion, you may find yourself less interested in them and their care. If they don’t like your opinion on news, they may start to question your ability as a doctor.
That’s not what we strive for, but it can be human nature. For better or worse we often reduce things to “us against them,” and learning someone is on the opposite side may, even subconsciously, alter how you treat them.
That’s not good, so to me it’s best not to know.
Some may think I’m being petty, or aloof, to be unwilling to discuss nonmedical issues of significance, but I don’t see it that way. Time is limited at the appointment and is best spent on medical care. Something unrelated to the visit that may alter my objective opinion of a patient – or theirs of me as a doctor – is best left out of it.
I’m here to be your doctor, and to do the best I can for you. I’m not here to be a debate partner. Whenever a patient asks me a question on politics or news I always think of the Monty Python skit “Argument Clinic.” That’s not why you’re here. There are plenty places to discuss such things. My office isn’t one of them.
Dr. Block has a solo neurology practice in Scottsdale, Ariz.
“What do you think about the whole Israel thing?”
That question came at the end of an otherwise routine appointment.
Maybe she was just chatting. Maybe she wanted something deeper. I have no idea. I just said, “I don’t discuss those things with patients.”
My answer surprised her, but she didn’t push it. She paid her copay, scheduled a follow-up for 3 months, and left.
As I’ve written before, I try to avoid all news except the local weather. The sad reality is that most of it is bad and there’s nothing I can really do about it. It only upsets me, which isn’t good for my mental health and blood pressure, and if I can’t change it, what’s the point of knowing? It falls under the serenity prayer.
Of course, some news stories are too big not to hear something. I pass TVs in the doctors lounge or coffee house, hear others talking as I stand in line for the elevator, or see blurbs go by when checking the weather. It’s not entirely unavoidable.
I’m not trivializing the Middle East. But, to me, it’s not part of the doctor-patient relationship any more than my political views are. You run the risk of driving a wedge between you and the person you’re caring for. If you don’t like their opinion, you may find yourself less interested in them and their care. If they don’t like your opinion on news, they may start to question your ability as a doctor.
That’s not what we strive for, but it can be human nature. For better or worse we often reduce things to “us against them,” and learning someone is on the opposite side may, even subconsciously, alter how you treat them.
That’s not good, so to me it’s best not to know.
Some may think I’m being petty, or aloof, to be unwilling to discuss nonmedical issues of significance, but I don’t see it that way. Time is limited at the appointment and is best spent on medical care. Something unrelated to the visit that may alter my objective opinion of a patient – or theirs of me as a doctor – is best left out of it.
I’m here to be your doctor, and to do the best I can for you. I’m not here to be a debate partner. Whenever a patient asks me a question on politics or news I always think of the Monty Python skit “Argument Clinic.” That’s not why you’re here. There are plenty places to discuss such things. My office isn’t one of them.
Dr. Block has a solo neurology practice in Scottsdale, Ariz.
AI in medicine has a major Cassandra problem
This transcript has been edited for clarity.
Today I’m going to talk to you about a study at the cutting edge of modern medicine, one that uses an artificial intelligence (AI) model to guide care. But before I do, I need to take you back to the late Bronze Age, to a city located on the coast of what is now Turkey.
Troy’s towering walls made it seem unassailable, but that would not stop the Achaeans and their fleet of black ships from making landfall, and, after a siege, destroying the city. The destruction of Troy, as told in the Iliad and the Aeneid, was foretold by Cassandra, the daughter of King Priam and Priestess of Troy.
Cassandra had been given the gift of prophecy by the god Apollo in exchange for her favors. But after the gift was bestowed, she rejected the bright god and, in his rage, he added a curse to her blessing: that no one would ever believe her prophecies.
Thus it was that when her brother Paris set off to Sparta to abduct Helen, she warned him that his actions would lead to the downfall of their great city. He, of course, ignored her.
And you know the rest of the story.
Why am I telling you the story of Cassandra of Troy when we’re supposed to be talking about AI in medicine? Because AI has a major Cassandra problem.
The recent history of AI, and particularly the subset of AI known as machine learning in medicine, has been characterized by an accuracy arms race.
The electronic health record allows for the collection of volumes of data orders of magnitude greater than what we have ever been able to collect before. And all that data can be crunched by various algorithms to make predictions about, well, anything – whether a patient will be transferred to the intensive care unit, whether a GI bleed will need an intervention, whether someone will die in the next year.
Studies in this area tend to rely on retrospective datasets, and as time has gone on, better algorithms and more data have led to better and better predictions. In some simpler cases, machine-learning models have achieved near-perfect accuracy – Cassandra-level accuracy – as in the reading of chest x-rays for pneumonia, for example.
But as Cassandra teaches us, even perfect prediction is useless if no one believes you, if they don’t change their behavior. And this is the central problem of AI in medicine today. Many people are focusing on accuracy of the prediction but have forgotten that high accuracy is just table stakes for an AI model to be useful. It has to not only be accurate, but its use also has to change outcomes for patients. We need to be able to save Troy.
The best way to determine whether an AI model will help patients is to treat a model like we treat a new medication and evaluate it through a randomized trial. That’s what researchers, led by Shannon Walker of Vanderbilt University, Nashville, Tenn., did in a paper appearing in JAMA Network Open.
The model in question was one that predicted venous thromboembolism – blood clots – in hospitalized children. The model took in a variety of data points from the health record: a history of blood clot, history of cancer, presence of a central line, a variety of lab values. And the predictive model was very good – maybe not Cassandra good, but it achieved an AUC of 0.90, which means it had very high accuracy.
But again, accuracy is just table stakes.
The authors deployed the model in the live health record and recorded the results. For half of the kids, that was all that happened; no one actually saw the predictions. For those randomized to the intervention, the hematology team would be notified when the risk for clot was calculated to be greater than 2.5%. The hematology team would then contact the primary team to discuss prophylactic anticoagulation.
This is an elegant approach.
Let’s start with those table stakes – accuracy. The predictions were, by and large, pretty accurate in this trial. Of the 135 kids who developed blood clots, 121 had been flagged by the model in advance. That’s about 90%. The model flagged about 10% of kids who didn’t get a blood clot as well, but that’s not entirely surprising since the threshold for flagging was a 2.5% risk.
Given that the model preidentified almost every kid who would go on to develop a blood clot, it would make sense that kids randomized to the intervention would do better; after all, Cassandra was calling out her warnings.
But those kids didn’t do better. The rate of blood clot was no different between the group that used the accurate prediction model and the group that did not.
Why? Why does the use of an accurate model not necessarily improve outcomes?
First of all, a warning must lead to some change in management. Indeed, the kids in the intervention group were more likely to receive anticoagulation, but barely so. There were lots of reasons for this: physician preference, imminent discharge, active bleeding, and so on.
But let’s take a look at the 77 kids in the intervention arm who developed blood clots, because I think this is an instructive analysis.
Six of them did not meet the 2.5% threshold criteria, a case where the model missed its mark. Again, accuracy is table stakes.
Of the remaining 71, only 16 got a recommendation from the hematologist to start anticoagulation. Why not more? Well, the model identified some of the high-risk kids on the weekend, and it seems that the study team did not contact treatment teams during that time. That may account for about 40% of these cases. The remainder had some contraindication to anticoagulation.
Most tellingly, of the 16 who did get a recommendation to start anticoagulation, the recommendation was followed in only seven patients.
This is the gap between accurate prediction and the ability to change outcomes for patients. A prediction is useless if it is wrong, for sure. But it’s also useless if you don’t tell anyone about it. It’s useless if you tell someone but they can’t do anything about it. And it’s useless if they could do something about it but choose not to.
That’s the gulf that these models need to cross at this point. So, the next time some slick company tells you how accurate their AI model is, ask them if accuracy is really the most important thing. If they say, “Well, yes, of course,” then tell them about Cassandra.
Dr. F. Perry Wilson is associate professor of medicine and public health and director of the Clinical and Translational Research Accelerator at Yale University, New Haven, Conn. He has disclosed no relevant financial relationships.
A version of this article appeared on Medscape.com.
This transcript has been edited for clarity.
Today I’m going to talk to you about a study at the cutting edge of modern medicine, one that uses an artificial intelligence (AI) model to guide care. But before I do, I need to take you back to the late Bronze Age, to a city located on the coast of what is now Turkey.
Troy’s towering walls made it seem unassailable, but that would not stop the Achaeans and their fleet of black ships from making landfall, and, after a siege, destroying the city. The destruction of Troy, as told in the Iliad and the Aeneid, was foretold by Cassandra, the daughter of King Priam and Priestess of Troy.
Cassandra had been given the gift of prophecy by the god Apollo in exchange for her favors. But after the gift was bestowed, she rejected the bright god and, in his rage, he added a curse to her blessing: that no one would ever believe her prophecies.
Thus it was that when her brother Paris set off to Sparta to abduct Helen, she warned him that his actions would lead to the downfall of their great city. He, of course, ignored her.
And you know the rest of the story.
Why am I telling you the story of Cassandra of Troy when we’re supposed to be talking about AI in medicine? Because AI has a major Cassandra problem.
The recent history of AI, and particularly the subset of AI known as machine learning in medicine, has been characterized by an accuracy arms race.
The electronic health record allows for the collection of volumes of data orders of magnitude greater than what we have ever been able to collect before. And all that data can be crunched by various algorithms to make predictions about, well, anything – whether a patient will be transferred to the intensive care unit, whether a GI bleed will need an intervention, whether someone will die in the next year.
Studies in this area tend to rely on retrospective datasets, and as time has gone on, better algorithms and more data have led to better and better predictions. In some simpler cases, machine-learning models have achieved near-perfect accuracy – Cassandra-level accuracy – as in the reading of chest x-rays for pneumonia, for example.
But as Cassandra teaches us, even perfect prediction is useless if no one believes you, if they don’t change their behavior. And this is the central problem of AI in medicine today. Many people are focusing on accuracy of the prediction but have forgotten that high accuracy is just table stakes for an AI model to be useful. It has to not only be accurate, but its use also has to change outcomes for patients. We need to be able to save Troy.
The best way to determine whether an AI model will help patients is to treat a model like we treat a new medication and evaluate it through a randomized trial. That’s what researchers, led by Shannon Walker of Vanderbilt University, Nashville, Tenn., did in a paper appearing in JAMA Network Open.
The model in question was one that predicted venous thromboembolism – blood clots – in hospitalized children. The model took in a variety of data points from the health record: a history of blood clot, history of cancer, presence of a central line, a variety of lab values. And the predictive model was very good – maybe not Cassandra good, but it achieved an AUC of 0.90, which means it had very high accuracy.
But again, accuracy is just table stakes.
The authors deployed the model in the live health record and recorded the results. For half of the kids, that was all that happened; no one actually saw the predictions. For those randomized to the intervention, the hematology team would be notified when the risk for clot was calculated to be greater than 2.5%. The hematology team would then contact the primary team to discuss prophylactic anticoagulation.
This is an elegant approach.
Let’s start with those table stakes – accuracy. The predictions were, by and large, pretty accurate in this trial. Of the 135 kids who developed blood clots, 121 had been flagged by the model in advance. That’s about 90%. The model flagged about 10% of kids who didn’t get a blood clot as well, but that’s not entirely surprising since the threshold for flagging was a 2.5% risk.
Given that the model preidentified almost every kid who would go on to develop a blood clot, it would make sense that kids randomized to the intervention would do better; after all, Cassandra was calling out her warnings.
But those kids didn’t do better. The rate of blood clot was no different between the group that used the accurate prediction model and the group that did not.
Why? Why does the use of an accurate model not necessarily improve outcomes?
First of all, a warning must lead to some change in management. Indeed, the kids in the intervention group were more likely to receive anticoagulation, but barely so. There were lots of reasons for this: physician preference, imminent discharge, active bleeding, and so on.
But let’s take a look at the 77 kids in the intervention arm who developed blood clots, because I think this is an instructive analysis.
Six of them did not meet the 2.5% threshold criteria, a case where the model missed its mark. Again, accuracy is table stakes.
Of the remaining 71, only 16 got a recommendation from the hematologist to start anticoagulation. Why not more? Well, the model identified some of the high-risk kids on the weekend, and it seems that the study team did not contact treatment teams during that time. That may account for about 40% of these cases. The remainder had some contraindication to anticoagulation.
Most tellingly, of the 16 who did get a recommendation to start anticoagulation, the recommendation was followed in only seven patients.
This is the gap between accurate prediction and the ability to change outcomes for patients. A prediction is useless if it is wrong, for sure. But it’s also useless if you don’t tell anyone about it. It’s useless if you tell someone but they can’t do anything about it. And it’s useless if they could do something about it but choose not to.
That’s the gulf that these models need to cross at this point. So, the next time some slick company tells you how accurate their AI model is, ask them if accuracy is really the most important thing. If they say, “Well, yes, of course,” then tell them about Cassandra.
Dr. F. Perry Wilson is associate professor of medicine and public health and director of the Clinical and Translational Research Accelerator at Yale University, New Haven, Conn. He has disclosed no relevant financial relationships.
A version of this article appeared on Medscape.com.
This transcript has been edited for clarity.
Today I’m going to talk to you about a study at the cutting edge of modern medicine, one that uses an artificial intelligence (AI) model to guide care. But before I do, I need to take you back to the late Bronze Age, to a city located on the coast of what is now Turkey.
Troy’s towering walls made it seem unassailable, but that would not stop the Achaeans and their fleet of black ships from making landfall, and, after a siege, destroying the city. The destruction of Troy, as told in the Iliad and the Aeneid, was foretold by Cassandra, the daughter of King Priam and Priestess of Troy.
Cassandra had been given the gift of prophecy by the god Apollo in exchange for her favors. But after the gift was bestowed, she rejected the bright god and, in his rage, he added a curse to her blessing: that no one would ever believe her prophecies.
Thus it was that when her brother Paris set off to Sparta to abduct Helen, she warned him that his actions would lead to the downfall of their great city. He, of course, ignored her.
And you know the rest of the story.
Why am I telling you the story of Cassandra of Troy when we’re supposed to be talking about AI in medicine? Because AI has a major Cassandra problem.
The recent history of AI, and particularly the subset of AI known as machine learning in medicine, has been characterized by an accuracy arms race.
The electronic health record allows for the collection of volumes of data orders of magnitude greater than what we have ever been able to collect before. And all that data can be crunched by various algorithms to make predictions about, well, anything – whether a patient will be transferred to the intensive care unit, whether a GI bleed will need an intervention, whether someone will die in the next year.
Studies in this area tend to rely on retrospective datasets, and as time has gone on, better algorithms and more data have led to better and better predictions. In some simpler cases, machine-learning models have achieved near-perfect accuracy – Cassandra-level accuracy – as in the reading of chest x-rays for pneumonia, for example.
But as Cassandra teaches us, even perfect prediction is useless if no one believes you, if they don’t change their behavior. And this is the central problem of AI in medicine today. Many people are focusing on accuracy of the prediction but have forgotten that high accuracy is just table stakes for an AI model to be useful. It has to not only be accurate, but its use also has to change outcomes for patients. We need to be able to save Troy.
The best way to determine whether an AI model will help patients is to treat a model like we treat a new medication and evaluate it through a randomized trial. That’s what researchers, led by Shannon Walker of Vanderbilt University, Nashville, Tenn., did in a paper appearing in JAMA Network Open.
The model in question was one that predicted venous thromboembolism – blood clots – in hospitalized children. The model took in a variety of data points from the health record: a history of blood clot, history of cancer, presence of a central line, a variety of lab values. And the predictive model was very good – maybe not Cassandra good, but it achieved an AUC of 0.90, which means it had very high accuracy.
But again, accuracy is just table stakes.
The authors deployed the model in the live health record and recorded the results. For half of the kids, that was all that happened; no one actually saw the predictions. For those randomized to the intervention, the hematology team would be notified when the risk for clot was calculated to be greater than 2.5%. The hematology team would then contact the primary team to discuss prophylactic anticoagulation.
This is an elegant approach.
Let’s start with those table stakes – accuracy. The predictions were, by and large, pretty accurate in this trial. Of the 135 kids who developed blood clots, 121 had been flagged by the model in advance. That’s about 90%. The model flagged about 10% of kids who didn’t get a blood clot as well, but that’s not entirely surprising since the threshold for flagging was a 2.5% risk.
Given that the model preidentified almost every kid who would go on to develop a blood clot, it would make sense that kids randomized to the intervention would do better; after all, Cassandra was calling out her warnings.
But those kids didn’t do better. The rate of blood clot was no different between the group that used the accurate prediction model and the group that did not.
Why? Why does the use of an accurate model not necessarily improve outcomes?
First of all, a warning must lead to some change in management. Indeed, the kids in the intervention group were more likely to receive anticoagulation, but barely so. There were lots of reasons for this: physician preference, imminent discharge, active bleeding, and so on.
But let’s take a look at the 77 kids in the intervention arm who developed blood clots, because I think this is an instructive analysis.
Six of them did not meet the 2.5% threshold criteria, a case where the model missed its mark. Again, accuracy is table stakes.
Of the remaining 71, only 16 got a recommendation from the hematologist to start anticoagulation. Why not more? Well, the model identified some of the high-risk kids on the weekend, and it seems that the study team did not contact treatment teams during that time. That may account for about 40% of these cases. The remainder had some contraindication to anticoagulation.
Most tellingly, of the 16 who did get a recommendation to start anticoagulation, the recommendation was followed in only seven patients.
This is the gap between accurate prediction and the ability to change outcomes for patients. A prediction is useless if it is wrong, for sure. But it’s also useless if you don’t tell anyone about it. It’s useless if you tell someone but they can’t do anything about it. And it’s useless if they could do something about it but choose not to.
That’s the gulf that these models need to cross at this point. So, the next time some slick company tells you how accurate their AI model is, ask them if accuracy is really the most important thing. If they say, “Well, yes, of course,” then tell them about Cassandra.
Dr. F. Perry Wilson is associate professor of medicine and public health and director of the Clinical and Translational Research Accelerator at Yale University, New Haven, Conn. He has disclosed no relevant financial relationships.
A version of this article appeared on Medscape.com.
Debate: Is lasting remission of type 2 diabetes feasible in the real-world setting?
The prospect of remission of type 2 diabetes (T2D) has captured the hearts and minds of many patients with T2D and health care professionals, including myself.
I have changed my narrative when supporting my patients with T2D. I used to say that T2D is a progressive condition, but considering seminal recent evidence like the DiRECT trial, I now say that T2D can be a progressive condition. Through significant weight loss, patients can reverse it and achieve remission of T2D. This has given my patients hope that their T2D is no longer an inexorable condition. And hope, of course, is a powerful enabler of change.
However,
I therefore relished the opportunity to attend a debate on this topic at the annual meeting of the European Association for the Study of Diabetes in Hamburg, Germany, between Roy Taylor, MD, principal investigator for the DiRECT study and professor of medicine and metabolism at the University of Newcastle, England, and Kamlesh Khunti, MD, PhD, professor of primary care diabetes at the University of Leicester, England.
Remarkable weight loss
Dr. Taylor powerfully recapitulated the initial results of the DiRECT study. T2D remission was achieved in 46% of participants who underwent a low-energy formula diet (around 850 calories daily) for 3-5 months. After 2 years’ follow-up, an impressive 36% of participants were still in remission. Dr. Taylor then discussed unpublished 5-year extension follow-up data of the DiRECT study. Average weight loss in the remaining intervention group was 6.1 kg. I echo Taylor’s sentiment that this finding is remarkable in the context of a dietary study.
Overall, 13% of participants were still in remission, and this cohort maintained an average weight loss of 8.9 kg. Dr. Taylor concluded that lasting remission of T2D is indeed feasible in a primary care setting.
Yet he acknowledged that although remission appears feasible in the longer term, it was not necessarily easy, or indeed possible, for everyone. He used a wonderful analogy about climbing Mount Everest: It is feasible, but not everyone can or wants to climb it. And even if you try, you might not reach the top.
This analogy perfectly encapsulates the challenges I have observed when my patients have striven for T2D remission. In my opinion, intensive weight management with a low-energy formula diet is not a panacea for T2D but another tool in our toolbox to offer patients.
He also described some “jaw-dropping” results regarding incidence of cancer: There were no cases of cancer in the intervention group during the 5-year period, but there were eight cases of cancer in the control group. The latter figure is consistent with published data for cancer incidence in patients with T2D and the body mass index (BMI) inclusion criteria for the DiRECT study (a BMI of 27-45 kg/m2). Obesity is an established risk factor for 13 types of cancer, and excess body fat entails an approximately 17% increased risk for cancer-specific mortality. This indeed is a powerful motivator to facilitate meaningful lifestyle change.
In primary care, we also need to be aware that most weight regain usually occurs secondary to a life event (for example, financial, family, or illness). We should reiterate to our patients that weight regain is not a failure; it is just part of life. Once the life event has passed, rapid weight loss can be attempted again. In the “rescue plans” that were integral to the DiRECT study, participants were offered further periods of total diet replacement, depending on quantity of weight gain. In fact, 50% of participants in DiRECT required rescue therapy, and their outcomes, reassuringly, were the same as the other 50%.
Dr. Taylor also quoted data from the ReTUNE study suggesting that weight regain was less of an issue for those with initial BMI of 21-27, and there is “more bang for your buck” in approaching remission of T2D in patients with lower BMI. The fact that people with normal or near-normal BMI can also reverse their T2D was also a game changer for my clinical practice; the concept of an individual or personal fat threshold that results in T2D offers a pragmatic explanation to patients with T2D who are frustrated by the lack of improvements in cardiometabolic parameters despite significant weight loss.
Finally, Dr. Taylor acknowledged the breadth of the definition of T2D remission: A1c < 48 mmol/mol at least 2 months off all antidiabetic medication. This definition includes A1c values within the “prediabetes” range: 42-47 mmol/mol.
He cited 10-year cardiovascular risk data driven by hypertension and dyslipidemia before significant weight loss and compared it with 10-year cardiovascular risk data after significant weight loss. Cardiovascular risk profile was more favorable after weight loss, compared with controls with prediabetes without weight loss, even though some of the intervention group who lost significant weight still had an A1c of 42-47 mmol/mol. Dr. Taylor suggested that we not label these individuals who have lost significant weight as having prediabetes. Instead “postdiabetes” should be preferred, because these patients had more favorable cardiovascular profiles.
This is a very important take-home message for primary care: prediabetes is more than just dysglycemia.
New terminology proposed
Dr. Khunti outlined a recent large, systematic review that concluded that the definition of T2D remission encompassed substantial heterogeneity. This heterogeneity complicates the interpretation of previous research on T2D remission and complicates the implementation of remission pathways into routine clinical practice. Furthermore, Dr. Khunti highlighted a recent consensus report on the definition and interpretation of remission in T2D that explicitly stated that the underlying pathophysiology of T2D is rarely normalized completely by interventions, thus reducing the possibility of lasting remission.
Dr. Khunti also challenged the cardiovascular benefits seen after T2D remission. Recent Danish registry data were presented, demonstrating a twofold increased risk for major adverse cardiovascular events over 5 years in individuals who achieved remission of T2D, but not on glucose-lowering drug therapy.
Adherence to strict dietary interventions in the longer term was also addressed. Diet-induced weight loss causes changes in circulating hormones such as ghrelin, glucose-dependent insulinotropic polypeptide (GIP), and leptin, which mediate appetite and drive hunger and an increased preference for energy-dense foods (that is, high-fat or sugary foods), all of which encourage weight regain. Dr. Khunti suggested that other interventions, such as glucagon-like peptide 1 (GLP-1) receptor agonists or bariatric surgery, specifically target some of these hormonal responses.
The challenges in recruitment and retention for lifestyle studies were also discussed; they reflect the challenges of behavioral programs in primary care. The DiRECT study had 20% participation of screened candidates and an attrition rate approaching 30%. The seminal Diabetes Prevention Program study and Finnish Diabetes Prevention Study had similar results. At a population level, individuals do not appear to want to participate in behavioral programs.
Dr. Khunti also warned that the review of annual care processes for diabetes is declining for patients who had achieved remission, possibly because of a false sense of reassurance among health care professionals. It is essential that all those in remission remain under at least annual follow-up, because there is still a risk for future microvascular and macrovascular complications, especially in the event of weight regain.
Dr. Khunti concluded by proposing new terminology for remission: remission of hyperglycemia or euglycemia, aiming for A1c < 48 mmol/mol with or without glucose-lowering therapy. I do agree with this; it reflects the zeitgeist of cardiorenal protective diabetes therapies and is analogous to rheumatoid arthritis, where remission is defined as no disease activity while on therapy. But one size does not fit all.
Sir William Osler’s words provide a fitting conclusion: “If it were not for the great variability among individuals, medicine might as well be a science and not an art.”
Dr. Fernando has disclosed that he has received speakers’ fees from Eli Lilly and Novo Nordisk.
Dr. Fernando is a general practitioner near Edinburgh, with a specialist interest in diabetes; cardiovascular, renal, and metabolic diseases; and medical education.
A version of this article first appeared on Medscape.com.
The prospect of remission of type 2 diabetes (T2D) has captured the hearts and minds of many patients with T2D and health care professionals, including myself.
I have changed my narrative when supporting my patients with T2D. I used to say that T2D is a progressive condition, but considering seminal recent evidence like the DiRECT trial, I now say that T2D can be a progressive condition. Through significant weight loss, patients can reverse it and achieve remission of T2D. This has given my patients hope that their T2D is no longer an inexorable condition. And hope, of course, is a powerful enabler of change.
However,
I therefore relished the opportunity to attend a debate on this topic at the annual meeting of the European Association for the Study of Diabetes in Hamburg, Germany, between Roy Taylor, MD, principal investigator for the DiRECT study and professor of medicine and metabolism at the University of Newcastle, England, and Kamlesh Khunti, MD, PhD, professor of primary care diabetes at the University of Leicester, England.
Remarkable weight loss
Dr. Taylor powerfully recapitulated the initial results of the DiRECT study. T2D remission was achieved in 46% of participants who underwent a low-energy formula diet (around 850 calories daily) for 3-5 months. After 2 years’ follow-up, an impressive 36% of participants were still in remission. Dr. Taylor then discussed unpublished 5-year extension follow-up data of the DiRECT study. Average weight loss in the remaining intervention group was 6.1 kg. I echo Taylor’s sentiment that this finding is remarkable in the context of a dietary study.
Overall, 13% of participants were still in remission, and this cohort maintained an average weight loss of 8.9 kg. Dr. Taylor concluded that lasting remission of T2D is indeed feasible in a primary care setting.
Yet he acknowledged that although remission appears feasible in the longer term, it was not necessarily easy, or indeed possible, for everyone. He used a wonderful analogy about climbing Mount Everest: It is feasible, but not everyone can or wants to climb it. And even if you try, you might not reach the top.
This analogy perfectly encapsulates the challenges I have observed when my patients have striven for T2D remission. In my opinion, intensive weight management with a low-energy formula diet is not a panacea for T2D but another tool in our toolbox to offer patients.
He also described some “jaw-dropping” results regarding incidence of cancer: There were no cases of cancer in the intervention group during the 5-year period, but there were eight cases of cancer in the control group. The latter figure is consistent with published data for cancer incidence in patients with T2D and the body mass index (BMI) inclusion criteria for the DiRECT study (a BMI of 27-45 kg/m2). Obesity is an established risk factor for 13 types of cancer, and excess body fat entails an approximately 17% increased risk for cancer-specific mortality. This indeed is a powerful motivator to facilitate meaningful lifestyle change.
In primary care, we also need to be aware that most weight regain usually occurs secondary to a life event (for example, financial, family, or illness). We should reiterate to our patients that weight regain is not a failure; it is just part of life. Once the life event has passed, rapid weight loss can be attempted again. In the “rescue plans” that were integral to the DiRECT study, participants were offered further periods of total diet replacement, depending on quantity of weight gain. In fact, 50% of participants in DiRECT required rescue therapy, and their outcomes, reassuringly, were the same as the other 50%.
Dr. Taylor also quoted data from the ReTUNE study suggesting that weight regain was less of an issue for those with initial BMI of 21-27, and there is “more bang for your buck” in approaching remission of T2D in patients with lower BMI. The fact that people with normal or near-normal BMI can also reverse their T2D was also a game changer for my clinical practice; the concept of an individual or personal fat threshold that results in T2D offers a pragmatic explanation to patients with T2D who are frustrated by the lack of improvements in cardiometabolic parameters despite significant weight loss.
Finally, Dr. Taylor acknowledged the breadth of the definition of T2D remission: A1c < 48 mmol/mol at least 2 months off all antidiabetic medication. This definition includes A1c values within the “prediabetes” range: 42-47 mmol/mol.
He cited 10-year cardiovascular risk data driven by hypertension and dyslipidemia before significant weight loss and compared it with 10-year cardiovascular risk data after significant weight loss. Cardiovascular risk profile was more favorable after weight loss, compared with controls with prediabetes without weight loss, even though some of the intervention group who lost significant weight still had an A1c of 42-47 mmol/mol. Dr. Taylor suggested that we not label these individuals who have lost significant weight as having prediabetes. Instead “postdiabetes” should be preferred, because these patients had more favorable cardiovascular profiles.
This is a very important take-home message for primary care: prediabetes is more than just dysglycemia.
New terminology proposed
Dr. Khunti outlined a recent large, systematic review that concluded that the definition of T2D remission encompassed substantial heterogeneity. This heterogeneity complicates the interpretation of previous research on T2D remission and complicates the implementation of remission pathways into routine clinical practice. Furthermore, Dr. Khunti highlighted a recent consensus report on the definition and interpretation of remission in T2D that explicitly stated that the underlying pathophysiology of T2D is rarely normalized completely by interventions, thus reducing the possibility of lasting remission.
Dr. Khunti also challenged the cardiovascular benefits seen after T2D remission. Recent Danish registry data were presented, demonstrating a twofold increased risk for major adverse cardiovascular events over 5 years in individuals who achieved remission of T2D, but not on glucose-lowering drug therapy.
Adherence to strict dietary interventions in the longer term was also addressed. Diet-induced weight loss causes changes in circulating hormones such as ghrelin, glucose-dependent insulinotropic polypeptide (GIP), and leptin, which mediate appetite and drive hunger and an increased preference for energy-dense foods (that is, high-fat or sugary foods), all of which encourage weight regain. Dr. Khunti suggested that other interventions, such as glucagon-like peptide 1 (GLP-1) receptor agonists or bariatric surgery, specifically target some of these hormonal responses.
The challenges in recruitment and retention for lifestyle studies were also discussed; they reflect the challenges of behavioral programs in primary care. The DiRECT study had 20% participation of screened candidates and an attrition rate approaching 30%. The seminal Diabetes Prevention Program study and Finnish Diabetes Prevention Study had similar results. At a population level, individuals do not appear to want to participate in behavioral programs.
Dr. Khunti also warned that the review of annual care processes for diabetes is declining for patients who had achieved remission, possibly because of a false sense of reassurance among health care professionals. It is essential that all those in remission remain under at least annual follow-up, because there is still a risk for future microvascular and macrovascular complications, especially in the event of weight regain.
Dr. Khunti concluded by proposing new terminology for remission: remission of hyperglycemia or euglycemia, aiming for A1c < 48 mmol/mol with or without glucose-lowering therapy. I do agree with this; it reflects the zeitgeist of cardiorenal protective diabetes therapies and is analogous to rheumatoid arthritis, where remission is defined as no disease activity while on therapy. But one size does not fit all.
Sir William Osler’s words provide a fitting conclusion: “If it were not for the great variability among individuals, medicine might as well be a science and not an art.”
Dr. Fernando has disclosed that he has received speakers’ fees from Eli Lilly and Novo Nordisk.
Dr. Fernando is a general practitioner near Edinburgh, with a specialist interest in diabetes; cardiovascular, renal, and metabolic diseases; and medical education.
A version of this article first appeared on Medscape.com.
The prospect of remission of type 2 diabetes (T2D) has captured the hearts and minds of many patients with T2D and health care professionals, including myself.
I have changed my narrative when supporting my patients with T2D. I used to say that T2D is a progressive condition, but considering seminal recent evidence like the DiRECT trial, I now say that T2D can be a progressive condition. Through significant weight loss, patients can reverse it and achieve remission of T2D. This has given my patients hope that their T2D is no longer an inexorable condition. And hope, of course, is a powerful enabler of change.
However,
I therefore relished the opportunity to attend a debate on this topic at the annual meeting of the European Association for the Study of Diabetes in Hamburg, Germany, between Roy Taylor, MD, principal investigator for the DiRECT study and professor of medicine and metabolism at the University of Newcastle, England, and Kamlesh Khunti, MD, PhD, professor of primary care diabetes at the University of Leicester, England.
Remarkable weight loss
Dr. Taylor powerfully recapitulated the initial results of the DiRECT study. T2D remission was achieved in 46% of participants who underwent a low-energy formula diet (around 850 calories daily) for 3-5 months. After 2 years’ follow-up, an impressive 36% of participants were still in remission. Dr. Taylor then discussed unpublished 5-year extension follow-up data of the DiRECT study. Average weight loss in the remaining intervention group was 6.1 kg. I echo Taylor’s sentiment that this finding is remarkable in the context of a dietary study.
Overall, 13% of participants were still in remission, and this cohort maintained an average weight loss of 8.9 kg. Dr. Taylor concluded that lasting remission of T2D is indeed feasible in a primary care setting.
Yet he acknowledged that although remission appears feasible in the longer term, it was not necessarily easy, or indeed possible, for everyone. He used a wonderful analogy about climbing Mount Everest: It is feasible, but not everyone can or wants to climb it. And even if you try, you might not reach the top.
This analogy perfectly encapsulates the challenges I have observed when my patients have striven for T2D remission. In my opinion, intensive weight management with a low-energy formula diet is not a panacea for T2D but another tool in our toolbox to offer patients.
He also described some “jaw-dropping” results regarding incidence of cancer: There were no cases of cancer in the intervention group during the 5-year period, but there were eight cases of cancer in the control group. The latter figure is consistent with published data for cancer incidence in patients with T2D and the body mass index (BMI) inclusion criteria for the DiRECT study (a BMI of 27-45 kg/m2). Obesity is an established risk factor for 13 types of cancer, and excess body fat entails an approximately 17% increased risk for cancer-specific mortality. This indeed is a powerful motivator to facilitate meaningful lifestyle change.
In primary care, we also need to be aware that most weight regain usually occurs secondary to a life event (for example, financial, family, or illness). We should reiterate to our patients that weight regain is not a failure; it is just part of life. Once the life event has passed, rapid weight loss can be attempted again. In the “rescue plans” that were integral to the DiRECT study, participants were offered further periods of total diet replacement, depending on quantity of weight gain. In fact, 50% of participants in DiRECT required rescue therapy, and their outcomes, reassuringly, were the same as the other 50%.
Dr. Taylor also quoted data from the ReTUNE study suggesting that weight regain was less of an issue for those with initial BMI of 21-27, and there is “more bang for your buck” in approaching remission of T2D in patients with lower BMI. The fact that people with normal or near-normal BMI can also reverse their T2D was also a game changer for my clinical practice; the concept of an individual or personal fat threshold that results in T2D offers a pragmatic explanation to patients with T2D who are frustrated by the lack of improvements in cardiometabolic parameters despite significant weight loss.
Finally, Dr. Taylor acknowledged the breadth of the definition of T2D remission: A1c < 48 mmol/mol at least 2 months off all antidiabetic medication. This definition includes A1c values within the “prediabetes” range: 42-47 mmol/mol.
He cited 10-year cardiovascular risk data driven by hypertension and dyslipidemia before significant weight loss and compared it with 10-year cardiovascular risk data after significant weight loss. Cardiovascular risk profile was more favorable after weight loss, compared with controls with prediabetes without weight loss, even though some of the intervention group who lost significant weight still had an A1c of 42-47 mmol/mol. Dr. Taylor suggested that we not label these individuals who have lost significant weight as having prediabetes. Instead “postdiabetes” should be preferred, because these patients had more favorable cardiovascular profiles.
This is a very important take-home message for primary care: prediabetes is more than just dysglycemia.
New terminology proposed
Dr. Khunti outlined a recent large, systematic review that concluded that the definition of T2D remission encompassed substantial heterogeneity. This heterogeneity complicates the interpretation of previous research on T2D remission and complicates the implementation of remission pathways into routine clinical practice. Furthermore, Dr. Khunti highlighted a recent consensus report on the definition and interpretation of remission in T2D that explicitly stated that the underlying pathophysiology of T2D is rarely normalized completely by interventions, thus reducing the possibility of lasting remission.
Dr. Khunti also challenged the cardiovascular benefits seen after T2D remission. Recent Danish registry data were presented, demonstrating a twofold increased risk for major adverse cardiovascular events over 5 years in individuals who achieved remission of T2D, but not on glucose-lowering drug therapy.
Adherence to strict dietary interventions in the longer term was also addressed. Diet-induced weight loss causes changes in circulating hormones such as ghrelin, glucose-dependent insulinotropic polypeptide (GIP), and leptin, which mediate appetite and drive hunger and an increased preference for energy-dense foods (that is, high-fat or sugary foods), all of which encourage weight regain. Dr. Khunti suggested that other interventions, such as glucagon-like peptide 1 (GLP-1) receptor agonists or bariatric surgery, specifically target some of these hormonal responses.
The challenges in recruitment and retention for lifestyle studies were also discussed; they reflect the challenges of behavioral programs in primary care. The DiRECT study had 20% participation of screened candidates and an attrition rate approaching 30%. The seminal Diabetes Prevention Program study and Finnish Diabetes Prevention Study had similar results. At a population level, individuals do not appear to want to participate in behavioral programs.
Dr. Khunti also warned that the review of annual care processes for diabetes is declining for patients who had achieved remission, possibly because of a false sense of reassurance among health care professionals. It is essential that all those in remission remain under at least annual follow-up, because there is still a risk for future microvascular and macrovascular complications, especially in the event of weight regain.
Dr. Khunti concluded by proposing new terminology for remission: remission of hyperglycemia or euglycemia, aiming for A1c < 48 mmol/mol with or without glucose-lowering therapy. I do agree with this; it reflects the zeitgeist of cardiorenal protective diabetes therapies and is analogous to rheumatoid arthritis, where remission is defined as no disease activity while on therapy. But one size does not fit all.
Sir William Osler’s words provide a fitting conclusion: “If it were not for the great variability among individuals, medicine might as well be a science and not an art.”
Dr. Fernando has disclosed that he has received speakers’ fees from Eli Lilly and Novo Nordisk.
Dr. Fernando is a general practitioner near Edinburgh, with a specialist interest in diabetes; cardiovascular, renal, and metabolic diseases; and medical education.
A version of this article first appeared on Medscape.com.
Making time to care for patients with diabetes
Can busy primary care offices continue to care for patients with diabetes? No one would argue that it is involved and takes effort, and health care providers are bankrupt when it comes to sparing additional time for this chronic disease. With roughly 37 million people living with diabetes and 96 million with prediabetes or early type 2 diabetes, and just over 8,000 practicing endocrinologists in the United States, we all need to make time especially in primary care to provide insight and holistic care. With limited time and budget, how do we do this?
First, decide to be involved in caring for patients with diabetes. Diabetes is best managed by interprofessional care teams, so you’re not going it alone. These teams may include physicians; pharmacists; physician assistants; advanced practice nurses; registered nurses; certified diabetes care and education specialists (CDCES); dietitians; and other professionals such as social workers, behavioral health professionals, medical assistants, and community health workers. Know which professionals are available to serve on your team, either within your clinic or as a consultant, and reach out to them to share the care and ease the burden. Remember to refer to these professionals to reinforce the diabetes intervention message to the patient.
Second, incorporate “diabetes only” appointments into your schedule, allowing time to focus on current comprehensive diabetes treatment goals, barriers/inertia for care. Remember to have short-interval follow-up as needed to keep that patient engaged to achieve their targets. Instruct your office staff to create diabetes appointment templates and reminders to patients to bring diabetes-related technologies, medication lists, and diabetes questions to the appointment. When I implemented this change, my patients welcomed the focus on their diabetes health, and they knew we were prioritizing this disease that they have for a lifetime. These appointments did not take away from their other conditions; rather, they often reminded me to stay focused on their diabetes and associated coconditions.
Taking the time to establish efficient workflows before implementing diabetes care saves countless hours later and immediately maximizes health care provider–patient interactions. Assign specific staff duties and expectations related to diabetes appointments, such as downloading diabetes technology, medication reconciliation, laboratory data, point-of-care hemoglobin A1c, basic foot exam, and patient goals for diabetes care. This allows the prescriber to focus on the glycemic, cardiologic, renal, and metabolic goals and overcome the therapeutic inertia that plagues us all.
Incorporating diabetes-related technology into clinical practice can be a significant time-saver but requires initial onboarding. Set aside a few hours to create a technology clinic flow, and designate at least one team member to be responsible for obtaining patient data before, during, or after encounters. If possible, obtaining data ahead of the visit will enhance efficiency, allowing for meaningful discussion of blood glucose and lifestyle patterns. Diabetes technology reveals the gaps in care and enhances our ability to identify the areas where glycemic intervention is needed. In addition, it reveals the impact of food choices, activity level, stress, and medication adherence to the person living with diabetes.
Finally, be proactive about therapeutic inertia. This is defined as a prescribers’ failure to intensify or deintensify a patient’s treatment when appropriate to do so. Causes of therapeutic inertia can be placed at the primary care physician level, including time constraints or inexperience in treating diabetes; the patient level, such as concerns about side effects or new treatment regimens; or a systemic level, such as availability of medications or their costs. Be real with yourself: We all have inertia and can identify areas to overcome. Never let inertia be traced back to you.
Not all inertia lives with the health care provider. Patients bring apprehension and concerns, have questions, and just want to share the frustrations associated with living their best life with the disease. Don’t assume that you know what your patients’ treatment barriers are; ask them. If you don’t have an answer, then note it and come up with one by the next follow-up. Remember that this is a chronic disease – a marathon, not a sprint. You don’t have to solve everything at one appointment; rather, keep the momentum going.
Let’s put this into clinical practice. For the next patient with diabetes who comes into your office, discuss with them your intention to prioritize their diabetes by having an appointment set aside to specifically focus on their individual goals and targets for their disease. Have the patient list any barriers and treatment goals they would like to review; flag your schedule to indicate it is a diabetes-only visit; and orient your staff to reconcile diabetes medications and record the patient’s last eye exam, urine albumin-to-creatinine ratio, A1c result, and blood glucose data. During this encounter, identify the patient’s personal targets for control, examine their feet, and review or order necessary laboratory metrics. Explore the patient-reported barriers and make inroads to remove or alleviate these. Advance treatment intervention, and schedule follow-up: every 4-6 weeks if the A1c is > 9%, every 2 months if it’s 7% to < 9%, and every 3-6 months if it’s < 7%. Utilize team diabetes care, such as CDCES referrals, dietitians, online resources, and community members, to help reinforce care and enhance engagement.
We need to take steps in our clinical practice to make the necessary space to accommodate this pervasive disease affecting nearly one-third of our population. Take a moment to look up and determine what needs to be in place so that you can take care of the people in your practice with diabetes. Laying the groundwork for implementing diabetes-only appointments can be time-consuming, but establishing consistent procedures, developing efficient workflows, and clearly defining roles and responsibilities is well worth the effort. This solid foundation equips the office, health care providers, and staff to care for persons with diabetes and will be invaluable to ensure that time for this care is available in the day-to-day clinical practice.
A version of this article first appeared on Medscape.com.
Can busy primary care offices continue to care for patients with diabetes? No one would argue that it is involved and takes effort, and health care providers are bankrupt when it comes to sparing additional time for this chronic disease. With roughly 37 million people living with diabetes and 96 million with prediabetes or early type 2 diabetes, and just over 8,000 practicing endocrinologists in the United States, we all need to make time especially in primary care to provide insight and holistic care. With limited time and budget, how do we do this?
First, decide to be involved in caring for patients with diabetes. Diabetes is best managed by interprofessional care teams, so you’re not going it alone. These teams may include physicians; pharmacists; physician assistants; advanced practice nurses; registered nurses; certified diabetes care and education specialists (CDCES); dietitians; and other professionals such as social workers, behavioral health professionals, medical assistants, and community health workers. Know which professionals are available to serve on your team, either within your clinic or as a consultant, and reach out to them to share the care and ease the burden. Remember to refer to these professionals to reinforce the diabetes intervention message to the patient.
Second, incorporate “diabetes only” appointments into your schedule, allowing time to focus on current comprehensive diabetes treatment goals, barriers/inertia for care. Remember to have short-interval follow-up as needed to keep that patient engaged to achieve their targets. Instruct your office staff to create diabetes appointment templates and reminders to patients to bring diabetes-related technologies, medication lists, and diabetes questions to the appointment. When I implemented this change, my patients welcomed the focus on their diabetes health, and they knew we were prioritizing this disease that they have for a lifetime. These appointments did not take away from their other conditions; rather, they often reminded me to stay focused on their diabetes and associated coconditions.
Taking the time to establish efficient workflows before implementing diabetes care saves countless hours later and immediately maximizes health care provider–patient interactions. Assign specific staff duties and expectations related to diabetes appointments, such as downloading diabetes technology, medication reconciliation, laboratory data, point-of-care hemoglobin A1c, basic foot exam, and patient goals for diabetes care. This allows the prescriber to focus on the glycemic, cardiologic, renal, and metabolic goals and overcome the therapeutic inertia that plagues us all.
Incorporating diabetes-related technology into clinical practice can be a significant time-saver but requires initial onboarding. Set aside a few hours to create a technology clinic flow, and designate at least one team member to be responsible for obtaining patient data before, during, or after encounters. If possible, obtaining data ahead of the visit will enhance efficiency, allowing for meaningful discussion of blood glucose and lifestyle patterns. Diabetes technology reveals the gaps in care and enhances our ability to identify the areas where glycemic intervention is needed. In addition, it reveals the impact of food choices, activity level, stress, and medication adherence to the person living with diabetes.
Finally, be proactive about therapeutic inertia. This is defined as a prescribers’ failure to intensify or deintensify a patient’s treatment when appropriate to do so. Causes of therapeutic inertia can be placed at the primary care physician level, including time constraints or inexperience in treating diabetes; the patient level, such as concerns about side effects or new treatment regimens; or a systemic level, such as availability of medications or their costs. Be real with yourself: We all have inertia and can identify areas to overcome. Never let inertia be traced back to you.
Not all inertia lives with the health care provider. Patients bring apprehension and concerns, have questions, and just want to share the frustrations associated with living their best life with the disease. Don’t assume that you know what your patients’ treatment barriers are; ask them. If you don’t have an answer, then note it and come up with one by the next follow-up. Remember that this is a chronic disease – a marathon, not a sprint. You don’t have to solve everything at one appointment; rather, keep the momentum going.
Let’s put this into clinical practice. For the next patient with diabetes who comes into your office, discuss with them your intention to prioritize their diabetes by having an appointment set aside to specifically focus on their individual goals and targets for their disease. Have the patient list any barriers and treatment goals they would like to review; flag your schedule to indicate it is a diabetes-only visit; and orient your staff to reconcile diabetes medications and record the patient’s last eye exam, urine albumin-to-creatinine ratio, A1c result, and blood glucose data. During this encounter, identify the patient’s personal targets for control, examine their feet, and review or order necessary laboratory metrics. Explore the patient-reported barriers and make inroads to remove or alleviate these. Advance treatment intervention, and schedule follow-up: every 4-6 weeks if the A1c is > 9%, every 2 months if it’s 7% to < 9%, and every 3-6 months if it’s < 7%. Utilize team diabetes care, such as CDCES referrals, dietitians, online resources, and community members, to help reinforce care and enhance engagement.
We need to take steps in our clinical practice to make the necessary space to accommodate this pervasive disease affecting nearly one-third of our population. Take a moment to look up and determine what needs to be in place so that you can take care of the people in your practice with diabetes. Laying the groundwork for implementing diabetes-only appointments can be time-consuming, but establishing consistent procedures, developing efficient workflows, and clearly defining roles and responsibilities is well worth the effort. This solid foundation equips the office, health care providers, and staff to care for persons with diabetes and will be invaluable to ensure that time for this care is available in the day-to-day clinical practice.
A version of this article first appeared on Medscape.com.
Can busy primary care offices continue to care for patients with diabetes? No one would argue that it is involved and takes effort, and health care providers are bankrupt when it comes to sparing additional time for this chronic disease. With roughly 37 million people living with diabetes and 96 million with prediabetes or early type 2 diabetes, and just over 8,000 practicing endocrinologists in the United States, we all need to make time especially in primary care to provide insight and holistic care. With limited time and budget, how do we do this?
First, decide to be involved in caring for patients with diabetes. Diabetes is best managed by interprofessional care teams, so you’re not going it alone. These teams may include physicians; pharmacists; physician assistants; advanced practice nurses; registered nurses; certified diabetes care and education specialists (CDCES); dietitians; and other professionals such as social workers, behavioral health professionals, medical assistants, and community health workers. Know which professionals are available to serve on your team, either within your clinic or as a consultant, and reach out to them to share the care and ease the burden. Remember to refer to these professionals to reinforce the diabetes intervention message to the patient.
Second, incorporate “diabetes only” appointments into your schedule, allowing time to focus on current comprehensive diabetes treatment goals, barriers/inertia for care. Remember to have short-interval follow-up as needed to keep that patient engaged to achieve their targets. Instruct your office staff to create diabetes appointment templates and reminders to patients to bring diabetes-related technologies, medication lists, and diabetes questions to the appointment. When I implemented this change, my patients welcomed the focus on their diabetes health, and they knew we were prioritizing this disease that they have for a lifetime. These appointments did not take away from their other conditions; rather, they often reminded me to stay focused on their diabetes and associated coconditions.
Taking the time to establish efficient workflows before implementing diabetes care saves countless hours later and immediately maximizes health care provider–patient interactions. Assign specific staff duties and expectations related to diabetes appointments, such as downloading diabetes technology, medication reconciliation, laboratory data, point-of-care hemoglobin A1c, basic foot exam, and patient goals for diabetes care. This allows the prescriber to focus on the glycemic, cardiologic, renal, and metabolic goals and overcome the therapeutic inertia that plagues us all.
Incorporating diabetes-related technology into clinical practice can be a significant time-saver but requires initial onboarding. Set aside a few hours to create a technology clinic flow, and designate at least one team member to be responsible for obtaining patient data before, during, or after encounters. If possible, obtaining data ahead of the visit will enhance efficiency, allowing for meaningful discussion of blood glucose and lifestyle patterns. Diabetes technology reveals the gaps in care and enhances our ability to identify the areas where glycemic intervention is needed. In addition, it reveals the impact of food choices, activity level, stress, and medication adherence to the person living with diabetes.
Finally, be proactive about therapeutic inertia. This is defined as a prescribers’ failure to intensify or deintensify a patient’s treatment when appropriate to do so. Causes of therapeutic inertia can be placed at the primary care physician level, including time constraints or inexperience in treating diabetes; the patient level, such as concerns about side effects or new treatment regimens; or a systemic level, such as availability of medications or their costs. Be real with yourself: We all have inertia and can identify areas to overcome. Never let inertia be traced back to you.
Not all inertia lives with the health care provider. Patients bring apprehension and concerns, have questions, and just want to share the frustrations associated with living their best life with the disease. Don’t assume that you know what your patients’ treatment barriers are; ask them. If you don’t have an answer, then note it and come up with one by the next follow-up. Remember that this is a chronic disease – a marathon, not a sprint. You don’t have to solve everything at one appointment; rather, keep the momentum going.
Let’s put this into clinical practice. For the next patient with diabetes who comes into your office, discuss with them your intention to prioritize their diabetes by having an appointment set aside to specifically focus on their individual goals and targets for their disease. Have the patient list any barriers and treatment goals they would like to review; flag your schedule to indicate it is a diabetes-only visit; and orient your staff to reconcile diabetes medications and record the patient’s last eye exam, urine albumin-to-creatinine ratio, A1c result, and blood glucose data. During this encounter, identify the patient’s personal targets for control, examine their feet, and review or order necessary laboratory metrics. Explore the patient-reported barriers and make inroads to remove or alleviate these. Advance treatment intervention, and schedule follow-up: every 4-6 weeks if the A1c is > 9%, every 2 months if it’s 7% to < 9%, and every 3-6 months if it’s < 7%. Utilize team diabetes care, such as CDCES referrals, dietitians, online resources, and community members, to help reinforce care and enhance engagement.
We need to take steps in our clinical practice to make the necessary space to accommodate this pervasive disease affecting nearly one-third of our population. Take a moment to look up and determine what needs to be in place so that you can take care of the people in your practice with diabetes. Laying the groundwork for implementing diabetes-only appointments can be time-consuming, but establishing consistent procedures, developing efficient workflows, and clearly defining roles and responsibilities is well worth the effort. This solid foundation equips the office, health care providers, and staff to care for persons with diabetes and will be invaluable to ensure that time for this care is available in the day-to-day clinical practice.
A version of this article first appeared on Medscape.com.
Suits or joggers? A doctor’s dress code
Look at this guy – NFL Chargers jersey and shorts with a RVCA hat on backward. And next to him, a woman wearing her spin-class-Lulu gear. There’s also a guy sporting a 2016 San Diego Rock ‘n Roll Marathon Tee. And that young woman is actually wearing slippers. A visitor from the 1950s would be thunderstruck to see such casual wear on people waiting to board a plane. Photos from that era show men buttoned up in white shirt and tie and women wearing Chanel with hats and white gloves. This dramatic transformation from formal to unfussy wear cuts through all social situations, including in my office. As a new doc out of residency, I used to wear a tie and shoes that could hold a shine. Now I wear jogger scrubs and sneakers. Rather than be offended by the lack of formality though, patients seem to appreciate it. Should they?
At first glance this seems to be a modern phenomenon. The reasons for casual wear today are manifold: about one-third of people work from home, Millennials are taking over with their TikTok values and general irreverence, COVID made us all fat and lazy. Heck, even the U.S. Senate briefly abolished the requirement to wear suits on the Senate floor. But getting dressed up was never to signal that you are elite or superior to others. It’s the opposite. To get dressed is a signal that you are serving others, a tradition that is as old as society.
Think of Downton Abbey as an example. The servants were always required to be smartly dressed when working, whereas members of the family could be dressed up or not. It’s clear who is serving whom. This tradition lives today in the hospitality industry. When you mosey into the lobby of a luxury hotel in your Rainbow sandals you can expect everyone who greets you will be in finery, signaling that they put in effort to serve you. You’ll find the same for all staff at the Mayo Clinic in Rochester, Minn., which is no coincidence.
Suits used to be standard in medicine. In the 19th century, physicians wore formal black-tie when seeing patients. Unlike hospitality however, we had good reason to eschew the tradition: germs. Once we figured out that our pus-stained ties and jackets were doing harm, we switched to wearing sanitized uniforms. Casual wear for doctors isn’t a modern phenomenon after all, then. For proof, compare Thomas Eakins painting “The Gross Clinic” (1875) with his later “The Agnew Clinic” (1889). In the former, Dr. Gross is portrayed in formal black wear, bloody hand and all. In the latter, Dr. Agnew is wearing white FIGS (or the 1890’s equivalent anyway). Similarly, nurses uniforms traditionally resembled kitchen servants, with criss-cross aprons and floor length skirts. It wasn’t until the 1980’s that nurses stopped wearing dresses and white caps.
In the operating theater it’s obviously critical that we wear sanitized scrubs to mitigate the risk of infection. Originally white to signal cleanliness, scrubs were changed to blue-green because surgeons were blinded by the lights bouncing off the uniforms. (Green is also opposite red on the color wheel, supposedly enhancing the ability to distinguish shades of red).
But Over time we’ve lost significant autonomy in our practice and lost a little respect from our patients. Payers tell us what to do. Patients question our expertise. Choosing what we wear is one of the few bits of medicine we still have agency. Pewter or pink, joggers or cargo pants, we get to choose.
The last time I flew British Airways everyone was in lounge wear, except the flight crew, of course. They were all smartly dressed. Recently British Airways rolled out updated, slightly more relaxed dress codes. Very modern, but I wonder if in a way we’re not all just a bit worse off.
Dr. Benabio is director of Healthcare Transformation and chief of dermatology at Kaiser Permanente San Diego. The opinions expressed in this column are his own and do not represent those of Kaiser Permanente. Dr. Benabio is @Dermdoc on Twitter. Write to him at [email protected]
Look at this guy – NFL Chargers jersey and shorts with a RVCA hat on backward. And next to him, a woman wearing her spin-class-Lulu gear. There’s also a guy sporting a 2016 San Diego Rock ‘n Roll Marathon Tee. And that young woman is actually wearing slippers. A visitor from the 1950s would be thunderstruck to see such casual wear on people waiting to board a plane. Photos from that era show men buttoned up in white shirt and tie and women wearing Chanel with hats and white gloves. This dramatic transformation from formal to unfussy wear cuts through all social situations, including in my office. As a new doc out of residency, I used to wear a tie and shoes that could hold a shine. Now I wear jogger scrubs and sneakers. Rather than be offended by the lack of formality though, patients seem to appreciate it. Should they?
At first glance this seems to be a modern phenomenon. The reasons for casual wear today are manifold: about one-third of people work from home, Millennials are taking over with their TikTok values and general irreverence, COVID made us all fat and lazy. Heck, even the U.S. Senate briefly abolished the requirement to wear suits on the Senate floor. But getting dressed up was never to signal that you are elite or superior to others. It’s the opposite. To get dressed is a signal that you are serving others, a tradition that is as old as society.
Think of Downton Abbey as an example. The servants were always required to be smartly dressed when working, whereas members of the family could be dressed up or not. It’s clear who is serving whom. This tradition lives today in the hospitality industry. When you mosey into the lobby of a luxury hotel in your Rainbow sandals you can expect everyone who greets you will be in finery, signaling that they put in effort to serve you. You’ll find the same for all staff at the Mayo Clinic in Rochester, Minn., which is no coincidence.
Suits used to be standard in medicine. In the 19th century, physicians wore formal black-tie when seeing patients. Unlike hospitality however, we had good reason to eschew the tradition: germs. Once we figured out that our pus-stained ties and jackets were doing harm, we switched to wearing sanitized uniforms. Casual wear for doctors isn’t a modern phenomenon after all, then. For proof, compare Thomas Eakins painting “The Gross Clinic” (1875) with his later “The Agnew Clinic” (1889). In the former, Dr. Gross is portrayed in formal black wear, bloody hand and all. In the latter, Dr. Agnew is wearing white FIGS (or the 1890’s equivalent anyway). Similarly, nurses uniforms traditionally resembled kitchen servants, with criss-cross aprons and floor length skirts. It wasn’t until the 1980’s that nurses stopped wearing dresses and white caps.
In the operating theater it’s obviously critical that we wear sanitized scrubs to mitigate the risk of infection. Originally white to signal cleanliness, scrubs were changed to blue-green because surgeons were blinded by the lights bouncing off the uniforms. (Green is also opposite red on the color wheel, supposedly enhancing the ability to distinguish shades of red).
But Over time we’ve lost significant autonomy in our practice and lost a little respect from our patients. Payers tell us what to do. Patients question our expertise. Choosing what we wear is one of the few bits of medicine we still have agency. Pewter or pink, joggers or cargo pants, we get to choose.
The last time I flew British Airways everyone was in lounge wear, except the flight crew, of course. They were all smartly dressed. Recently British Airways rolled out updated, slightly more relaxed dress codes. Very modern, but I wonder if in a way we’re not all just a bit worse off.
Dr. Benabio is director of Healthcare Transformation and chief of dermatology at Kaiser Permanente San Diego. The opinions expressed in this column are his own and do not represent those of Kaiser Permanente. Dr. Benabio is @Dermdoc on Twitter. Write to him at [email protected]
Look at this guy – NFL Chargers jersey and shorts with a RVCA hat on backward. And next to him, a woman wearing her spin-class-Lulu gear. There’s also a guy sporting a 2016 San Diego Rock ‘n Roll Marathon Tee. And that young woman is actually wearing slippers. A visitor from the 1950s would be thunderstruck to see such casual wear on people waiting to board a plane. Photos from that era show men buttoned up in white shirt and tie and women wearing Chanel with hats and white gloves. This dramatic transformation from formal to unfussy wear cuts through all social situations, including in my office. As a new doc out of residency, I used to wear a tie and shoes that could hold a shine. Now I wear jogger scrubs and sneakers. Rather than be offended by the lack of formality though, patients seem to appreciate it. Should they?
At first glance this seems to be a modern phenomenon. The reasons for casual wear today are manifold: about one-third of people work from home, Millennials are taking over with their TikTok values and general irreverence, COVID made us all fat and lazy. Heck, even the U.S. Senate briefly abolished the requirement to wear suits on the Senate floor. But getting dressed up was never to signal that you are elite or superior to others. It’s the opposite. To get dressed is a signal that you are serving others, a tradition that is as old as society.
Think of Downton Abbey as an example. The servants were always required to be smartly dressed when working, whereas members of the family could be dressed up or not. It’s clear who is serving whom. This tradition lives today in the hospitality industry. When you mosey into the lobby of a luxury hotel in your Rainbow sandals you can expect everyone who greets you will be in finery, signaling that they put in effort to serve you. You’ll find the same for all staff at the Mayo Clinic in Rochester, Minn., which is no coincidence.
Suits used to be standard in medicine. In the 19th century, physicians wore formal black-tie when seeing patients. Unlike hospitality however, we had good reason to eschew the tradition: germs. Once we figured out that our pus-stained ties and jackets were doing harm, we switched to wearing sanitized uniforms. Casual wear for doctors isn’t a modern phenomenon after all, then. For proof, compare Thomas Eakins painting “The Gross Clinic” (1875) with his later “The Agnew Clinic” (1889). In the former, Dr. Gross is portrayed in formal black wear, bloody hand and all. In the latter, Dr. Agnew is wearing white FIGS (or the 1890’s equivalent anyway). Similarly, nurses uniforms traditionally resembled kitchen servants, with criss-cross aprons and floor length skirts. It wasn’t until the 1980’s that nurses stopped wearing dresses and white caps.
In the operating theater it’s obviously critical that we wear sanitized scrubs to mitigate the risk of infection. Originally white to signal cleanliness, scrubs were changed to blue-green because surgeons were blinded by the lights bouncing off the uniforms. (Green is also opposite red on the color wheel, supposedly enhancing the ability to distinguish shades of red).
But Over time we’ve lost significant autonomy in our practice and lost a little respect from our patients. Payers tell us what to do. Patients question our expertise. Choosing what we wear is one of the few bits of medicine we still have agency. Pewter or pink, joggers or cargo pants, we get to choose.
The last time I flew British Airways everyone was in lounge wear, except the flight crew, of course. They were all smartly dressed. Recently British Airways rolled out updated, slightly more relaxed dress codes. Very modern, but I wonder if in a way we’re not all just a bit worse off.
Dr. Benabio is director of Healthcare Transformation and chief of dermatology at Kaiser Permanente San Diego. The opinions expressed in this column are his own and do not represent those of Kaiser Permanente. Dr. Benabio is @Dermdoc on Twitter. Write to him at [email protected]
Trading one’s eggs for a service discount raises tough issues, says ethicist
This transcript has been edited for clarity.
I had a case come to me of a 32-year-old resident who works in a hospital near where I am and was very interested in freezing her eggs. She wasn’t married and was getting worried that maybe she wouldn’t have a partner soon. She was also getting worried that the potential ability of her eggs to be fertilized would begin to decline, which is a phenomenon that does occur with age. She thought, I’m 32; maybe I should freeze my eggs now, as it’s better than to try freezing them when I’m 35 or 37. The potency may be far less.
There are many programs out there now. There have been academic programs for a long time that have been doing egg freezing, and there are many children who have been born successfully. However, it’s also true that people freeze their eggs when they’re 40 years old, and the likelihood of their “working,” if you will, is far less. I wouldn’t say it’s impossible, but age matters. This medical resident knew that and she decided to look into egg freezing.
Well, it turned out that egg freezing is not something that her student insurance plan – or most insurance plans in general – covers. The opportunity to do this is probably going to cost her about $10,000. There are many new egg-freezing infertility programs that have stared up that aren’t part of hospitals. There are clinics that are run for profit. They sometimes encourage women to freeze their eggs.
The student resident quickly found out that there were companies near her who would do egg freezing but would cut a deal if she agreed to take drugs to super-ovulate, make a large number of eggs, and they would be procured if she agreed to give half of them to other women who needed eggs for their infertility treatment. She could keep half and she could get very discounted treatment of egg freezing.
That may be a deal that she’s going to accept. She doesn’t have a path forward. She’s worried about freezing her eggs right now. But there are many ethical considerations that really have to be thought through here.
First and foremost, she’s giving eggs to others. They’re going to use them to try to make children. They can’t make their own eggs, for some reason. She’s going to have some biologically related kids out there. It used to be that you could say to someone who donated sperm or eggs that this will be anonymous.
But in today’s day and age with 23andMe, Ancestry, and better genetic testing, there’s a pretty good likelihood that somebody is going to find out that the person they thought was their biological mom isn’t, and they have someone out there who was the person who, in this case, donated an egg.
Is she willing to risk having that connection, that contact, to have someone enter her life in the future? It’s a situation where she’s donating the eggs, but I’ll tell you that the clinic is going to make far more money using the donated eggs, probably getting $10,000 or $15,000 a cycle with people who are trying to have a child. They’ll make much more money than she’s going to get by donating.
She may get a $5,000 discount, if you will, but the clinic has a business interest. The more they get women involved in bartering their eggs, the more they’re going to profit. In a sense, she’s being coerced, perhaps – I’m going to put it glibly – to sell cheaply. She probably is getting undervalue, even though she needs a path to do this egg freezing.
The other big issue is that we don’t know that egg freezing is going to work for her until someone tries to use those eggs. She may have her own infertility problem not due to age but to other things. Approximately 8%-9% of couples do have infertility problems, sometimes related to gametes. She may never get a partner. Maybe she doesn’t want to use these eggs on her own as a single mom. All of these issues have to be talked through.
What really troubles me here is not so much that someone would choose to barter their eggs, but that they don’t get counseling. They don’t get independent advice about thinking this all through. It’s turning into a business. A business has a commodity – her eggs – that they want. She’s getting more and more desperate, willing to cut a deal to get where she needs to be, but perhaps is not really thinking through all of the ethical dimensions that bartering or trading one’s eggs in order to gain access to freezing entails.
We have to set up a system where there’s independent advice and independent counseling; otherwise, I think we’re closer to exploitation.
A version of this article first appeared on Medscape.com.
Dr. Caplan is director, division of medical ethics, New York University Langone Medical Center, New York. He has served as a director, officer, partner, employee, advisor, consultant, or trustee for Johnson & Johnson’s Panel for Compassionate Drug Use.
This transcript has been edited for clarity.
I had a case come to me of a 32-year-old resident who works in a hospital near where I am and was very interested in freezing her eggs. She wasn’t married and was getting worried that maybe she wouldn’t have a partner soon. She was also getting worried that the potential ability of her eggs to be fertilized would begin to decline, which is a phenomenon that does occur with age. She thought, I’m 32; maybe I should freeze my eggs now, as it’s better than to try freezing them when I’m 35 or 37. The potency may be far less.
There are many programs out there now. There have been academic programs for a long time that have been doing egg freezing, and there are many children who have been born successfully. However, it’s also true that people freeze their eggs when they’re 40 years old, and the likelihood of their “working,” if you will, is far less. I wouldn’t say it’s impossible, but age matters. This medical resident knew that and she decided to look into egg freezing.
Well, it turned out that egg freezing is not something that her student insurance plan – or most insurance plans in general – covers. The opportunity to do this is probably going to cost her about $10,000. There are many new egg-freezing infertility programs that have stared up that aren’t part of hospitals. There are clinics that are run for profit. They sometimes encourage women to freeze their eggs.
The student resident quickly found out that there were companies near her who would do egg freezing but would cut a deal if she agreed to take drugs to super-ovulate, make a large number of eggs, and they would be procured if she agreed to give half of them to other women who needed eggs for their infertility treatment. She could keep half and she could get very discounted treatment of egg freezing.
That may be a deal that she’s going to accept. She doesn’t have a path forward. She’s worried about freezing her eggs right now. But there are many ethical considerations that really have to be thought through here.
First and foremost, she’s giving eggs to others. They’re going to use them to try to make children. They can’t make their own eggs, for some reason. She’s going to have some biologically related kids out there. It used to be that you could say to someone who donated sperm or eggs that this will be anonymous.
But in today’s day and age with 23andMe, Ancestry, and better genetic testing, there’s a pretty good likelihood that somebody is going to find out that the person they thought was their biological mom isn’t, and they have someone out there who was the person who, in this case, donated an egg.
Is she willing to risk having that connection, that contact, to have someone enter her life in the future? It’s a situation where she’s donating the eggs, but I’ll tell you that the clinic is going to make far more money using the donated eggs, probably getting $10,000 or $15,000 a cycle with people who are trying to have a child. They’ll make much more money than she’s going to get by donating.
She may get a $5,000 discount, if you will, but the clinic has a business interest. The more they get women involved in bartering their eggs, the more they’re going to profit. In a sense, she’s being coerced, perhaps – I’m going to put it glibly – to sell cheaply. She probably is getting undervalue, even though she needs a path to do this egg freezing.
The other big issue is that we don’t know that egg freezing is going to work for her until someone tries to use those eggs. She may have her own infertility problem not due to age but to other things. Approximately 8%-9% of couples do have infertility problems, sometimes related to gametes. She may never get a partner. Maybe she doesn’t want to use these eggs on her own as a single mom. All of these issues have to be talked through.
What really troubles me here is not so much that someone would choose to barter their eggs, but that they don’t get counseling. They don’t get independent advice about thinking this all through. It’s turning into a business. A business has a commodity – her eggs – that they want. She’s getting more and more desperate, willing to cut a deal to get where she needs to be, but perhaps is not really thinking through all of the ethical dimensions that bartering or trading one’s eggs in order to gain access to freezing entails.
We have to set up a system where there’s independent advice and independent counseling; otherwise, I think we’re closer to exploitation.
A version of this article first appeared on Medscape.com.
Dr. Caplan is director, division of medical ethics, New York University Langone Medical Center, New York. He has served as a director, officer, partner, employee, advisor, consultant, or trustee for Johnson & Johnson’s Panel for Compassionate Drug Use.
This transcript has been edited for clarity.
I had a case come to me of a 32-year-old resident who works in a hospital near where I am and was very interested in freezing her eggs. She wasn’t married and was getting worried that maybe she wouldn’t have a partner soon. She was also getting worried that the potential ability of her eggs to be fertilized would begin to decline, which is a phenomenon that does occur with age. She thought, I’m 32; maybe I should freeze my eggs now, as it’s better than to try freezing them when I’m 35 or 37. The potency may be far less.
There are many programs out there now. There have been academic programs for a long time that have been doing egg freezing, and there are many children who have been born successfully. However, it’s also true that people freeze their eggs when they’re 40 years old, and the likelihood of their “working,” if you will, is far less. I wouldn’t say it’s impossible, but age matters. This medical resident knew that and she decided to look into egg freezing.
Well, it turned out that egg freezing is not something that her student insurance plan – or most insurance plans in general – covers. The opportunity to do this is probably going to cost her about $10,000. There are many new egg-freezing infertility programs that have stared up that aren’t part of hospitals. There are clinics that are run for profit. They sometimes encourage women to freeze their eggs.
The student resident quickly found out that there were companies near her who would do egg freezing but would cut a deal if she agreed to take drugs to super-ovulate, make a large number of eggs, and they would be procured if she agreed to give half of them to other women who needed eggs for their infertility treatment. She could keep half and she could get very discounted treatment of egg freezing.
That may be a deal that she’s going to accept. She doesn’t have a path forward. She’s worried about freezing her eggs right now. But there are many ethical considerations that really have to be thought through here.
First and foremost, she’s giving eggs to others. They’re going to use them to try to make children. They can’t make their own eggs, for some reason. She’s going to have some biologically related kids out there. It used to be that you could say to someone who donated sperm or eggs that this will be anonymous.
But in today’s day and age with 23andMe, Ancestry, and better genetic testing, there’s a pretty good likelihood that somebody is going to find out that the person they thought was their biological mom isn’t, and they have someone out there who was the person who, in this case, donated an egg.
Is she willing to risk having that connection, that contact, to have someone enter her life in the future? It’s a situation where she’s donating the eggs, but I’ll tell you that the clinic is going to make far more money using the donated eggs, probably getting $10,000 or $15,000 a cycle with people who are trying to have a child. They’ll make much more money than she’s going to get by donating.
She may get a $5,000 discount, if you will, but the clinic has a business interest. The more they get women involved in bartering their eggs, the more they’re going to profit. In a sense, she’s being coerced, perhaps – I’m going to put it glibly – to sell cheaply. She probably is getting undervalue, even though she needs a path to do this egg freezing.
The other big issue is that we don’t know that egg freezing is going to work for her until someone tries to use those eggs. She may have her own infertility problem not due to age but to other things. Approximately 8%-9% of couples do have infertility problems, sometimes related to gametes. She may never get a partner. Maybe she doesn’t want to use these eggs on her own as a single mom. All of these issues have to be talked through.
What really troubles me here is not so much that someone would choose to barter their eggs, but that they don’t get counseling. They don’t get independent advice about thinking this all through. It’s turning into a business. A business has a commodity – her eggs – that they want. She’s getting more and more desperate, willing to cut a deal to get where she needs to be, but perhaps is not really thinking through all of the ethical dimensions that bartering or trading one’s eggs in order to gain access to freezing entails.
We have to set up a system where there’s independent advice and independent counseling; otherwise, I think we’re closer to exploitation.
A version of this article first appeared on Medscape.com.
Dr. Caplan is director, division of medical ethics, New York University Langone Medical Center, New York. He has served as a director, officer, partner, employee, advisor, consultant, or trustee for Johnson & Johnson’s Panel for Compassionate Drug Use.
How PCPs are penalized for positive outcomes from lifestyle change
The Centers for Medicare & Medicaid Services 2022 National Quality Strategy is described as an “ambitious long-term initiative that aims to promote the highest quality outcomes and safest care for all individuals.” It is a commendable goal for an overburdened U.S. health care system that spends more than other high-income counties yet experiences poorer outcomes. But whole-person, person-centered care cannot be achieved under current misaligned quality measures that fail to measure what we purport to value: the quintuple aim of improved health outcomes, cost savings, patient satisfaction, clinician well-being, and health equity.
Lifestyle first
Clinical practice guidelines for many chronic diseases recommend lifestyle intervention as the first and optimal treatment. A growing body of evidence supports lifestyle behavior interventions to treat and, when used intensively, even reverse common chronic conditions such as cardiovascular disease, obesity, and type 2 diabetes, while also providing effective prevention for those conditions. However, no current quality measures consider lifestyle interventions. In fact, some quality measures unintentionally penalize physicians for successfully treating or reversing disease through lifestyle behavior interventions while rewarding clinicians for meeting process measures – usually adherence to medication – regardless of whether health outcomes improved.
Rewarding medication adherence for the treatment of diseases in which lifestyle is a primary therapy (such as hypertension), combined with other health care constraints (lack of lifestyle education, time to spend with patients, and infrastructure support) incentivizes physicians to skip the conversation about lifestyle changes and go straight to medication prescription. Meanwhile, the clinician who takes the extra time to guide a patient toward lifestyle interventions that could treat their current disease and prevent future diseases – without side effects – is penalized.
Misaligned quality measures like these can stifle clinical judgment and risk reducing the practice of medicine to mindless box-checking. In many cases, patients are not even informed that lifestyle behavior change may be a treatment option (much less the first recommended option) for their conditions. This delivery of care is not person-centered and, in fact, may raise questions about the adequacy of informed treatment consent.
Reimbursement barriers
Lifestyle medicine is a growing medical specialty that uses therapeutic lifestyle interventions as a primary modality to treat chronic conditions. Since certification began in 2017, almost 2500 US physicians and 1000 nonphysician health professionals have earned certification. Health systems, including the U.S. military, are increasingly integrating lifestyle medicine. There have been advancements since one survey found that more than half of lifestyle medicine clinicians reported receiving no reimbursement for lifestyle behavior interventions. However, barriers, especially in fee-for-service systems, still inhibit many patients from receiving insurance coverage for comprehensive, interdisciplinary, and whole-person treatments called intensive therapeutic lifestyle change (ITLC) programs.
Existing comprehensive lifestyle programs that patients are eligible for (ie, the Diabetes Prevention Program and intensive behavioral therapy) are often so poorly reimbursed that clinicians and health systems decline to offer them. An example of a well-reimbursed ITLC program is intensive cardiac rehabilitation (ICR), which remains underutilized and limited to a narrow segment of patients, despite ICR›s proven benefits for managing comorbid risk factors such as hemoglobin A1c and weight. Even when lifestyle intervention programs are available and patients are eligible to participate (often through shared medical appointments), patient copays for the frequent visits required to achieve and sustain behavior change – or the lack of reimbursement for interdisciplinary team members – discourage engagement.
Penalizing successful outcomes
Despite the fact that lifestyle behaviors are top contributors to health and, conversely, contribute to up to 80% of chronic diseases, few quality measures focus on screening for lifestyle factors or treating diseases with lifestyle interventions. An example of an existing quality measure is screening or treatment for harmful substance use.
Specific quality measures that penalize lifestyle medicine approaches include pharmacotherapy for type 2 diabetes, dyslipidemia, osteoporosis, and gout as well as approaches to rheumatoid arthritis.
Statins offer a useful example of the conundrum faced by clinicians who want to offer lifestyle interventions. A lifestyle medicine primary care physician had a patient covered by Medicare Advantage who was diagnosed with hyperlipidemia. The patient had total cholesterol of 226 and a triglycerides level of 132. Instead of prescribing the routine statin, the physician prescribed lifestyle behavior modifications. Within 3 weeks, the patient›s total cholesterol improved to 171 and triglycerides to 75. This was a great success for the delighted patient. However, the CMS 5-Star Rating System assigned the primary care physician a grade of C rather than A, which put the physician›s 5-star rating at risk. Why? Because the system bases its score largely on medication compliance. The physician was penalized despite achieving the optimal health outcome, and at a lower cost than with medication. This misalignment does not incentivize patient-centered care because it disregards patient preference, shared decision-making, and evidence-based practice.
Risk adjustment
Rather than automatically managing disease with ever-increasing quantities of costly medications and procedures, lifestyle medicine clinicians first pursue a goal of health restoration when appropriate. But Medicare risk adjustment incentivizes physicians to manage rather than reverse disease. How much Medicare pays health plans is determined in part by how sick the patients are; the sicker the patient, the more Medicare pays, because those patients› costs are expected to be higher. This ensures that health plans are not penalized for enrolling sicker patients. But a physician utilizing diet alone to achieve remission in a patient with type 2 diabetes is penalized financially because, when the risk is adjusted, diabetes is no longer listed among the patient›s conditions. So, Medicare pays the physician less money. That misalignment incentivizes clinicians to manage the symptoms of type 2 diabetes rather than achieve remission, despite remission being the ideal clinical outcome.
Realigning quality measures
Quality measures were developed to quantify health care processes and outcomes, and to ensure the delivery of safe care to all patients. However, over time the number of quality measures has swelled to 2500, evolving into a confusing, time-consuming, and even soul-crushing responsibility for the physician.
Instead of relying heavily on process measures, we must incentivize outcome measures that honor patient autonomy and allow clinicians to offer lifestyle intervention as the first line of treatment. Risk-score calculations should be adjusted so that we stop incentivizing disease management and penalizing disease reversal.
CMS’s proposed development of “a universal foundation” of quality measures is an opportunity to begin the realignment of quality measures and values. This foundation is intended to establish more consistent and meaningful measures, reduce clinician burnout by streamlining the reporting process, and advance health equity. For this change to be successful, it is vital that lifestyle behavior interventions – optimal nutrition, physical activity, restorative sleep, social connections, stress management, and avoidance of harmful substances – become the foundation of universal quality measures. This will ensure that every clinician is incentivized to discuss lifestyle behaviors with patients and pursue the first clinical step recommended by clinical practice guidelines for most chronic diseases. Only then can we truly deliver high-value, whole-person, person-centered care and achieve the quintuple aim.
Dr. Patel is president-elect, American College of Lifestyle Medicine; Lifestyle Medicine Medical Director, Wellvana Health, Midland, Tex. She has disclosed no relevant financial relationships.
A version of this article first appeared on Medscape.com.
The Centers for Medicare & Medicaid Services 2022 National Quality Strategy is described as an “ambitious long-term initiative that aims to promote the highest quality outcomes and safest care for all individuals.” It is a commendable goal for an overburdened U.S. health care system that spends more than other high-income counties yet experiences poorer outcomes. But whole-person, person-centered care cannot be achieved under current misaligned quality measures that fail to measure what we purport to value: the quintuple aim of improved health outcomes, cost savings, patient satisfaction, clinician well-being, and health equity.
Lifestyle first
Clinical practice guidelines for many chronic diseases recommend lifestyle intervention as the first and optimal treatment. A growing body of evidence supports lifestyle behavior interventions to treat and, when used intensively, even reverse common chronic conditions such as cardiovascular disease, obesity, and type 2 diabetes, while also providing effective prevention for those conditions. However, no current quality measures consider lifestyle interventions. In fact, some quality measures unintentionally penalize physicians for successfully treating or reversing disease through lifestyle behavior interventions while rewarding clinicians for meeting process measures – usually adherence to medication – regardless of whether health outcomes improved.
Rewarding medication adherence for the treatment of diseases in which lifestyle is a primary therapy (such as hypertension), combined with other health care constraints (lack of lifestyle education, time to spend with patients, and infrastructure support) incentivizes physicians to skip the conversation about lifestyle changes and go straight to medication prescription. Meanwhile, the clinician who takes the extra time to guide a patient toward lifestyle interventions that could treat their current disease and prevent future diseases – without side effects – is penalized.
Misaligned quality measures like these can stifle clinical judgment and risk reducing the practice of medicine to mindless box-checking. In many cases, patients are not even informed that lifestyle behavior change may be a treatment option (much less the first recommended option) for their conditions. This delivery of care is not person-centered and, in fact, may raise questions about the adequacy of informed treatment consent.
Reimbursement barriers
Lifestyle medicine is a growing medical specialty that uses therapeutic lifestyle interventions as a primary modality to treat chronic conditions. Since certification began in 2017, almost 2500 US physicians and 1000 nonphysician health professionals have earned certification. Health systems, including the U.S. military, are increasingly integrating lifestyle medicine. There have been advancements since one survey found that more than half of lifestyle medicine clinicians reported receiving no reimbursement for lifestyle behavior interventions. However, barriers, especially in fee-for-service systems, still inhibit many patients from receiving insurance coverage for comprehensive, interdisciplinary, and whole-person treatments called intensive therapeutic lifestyle change (ITLC) programs.
Existing comprehensive lifestyle programs that patients are eligible for (ie, the Diabetes Prevention Program and intensive behavioral therapy) are often so poorly reimbursed that clinicians and health systems decline to offer them. An example of a well-reimbursed ITLC program is intensive cardiac rehabilitation (ICR), which remains underutilized and limited to a narrow segment of patients, despite ICR›s proven benefits for managing comorbid risk factors such as hemoglobin A1c and weight. Even when lifestyle intervention programs are available and patients are eligible to participate (often through shared medical appointments), patient copays for the frequent visits required to achieve and sustain behavior change – or the lack of reimbursement for interdisciplinary team members – discourage engagement.
Penalizing successful outcomes
Despite the fact that lifestyle behaviors are top contributors to health and, conversely, contribute to up to 80% of chronic diseases, few quality measures focus on screening for lifestyle factors or treating diseases with lifestyle interventions. An example of an existing quality measure is screening or treatment for harmful substance use.
Specific quality measures that penalize lifestyle medicine approaches include pharmacotherapy for type 2 diabetes, dyslipidemia, osteoporosis, and gout as well as approaches to rheumatoid arthritis.
Statins offer a useful example of the conundrum faced by clinicians who want to offer lifestyle interventions. A lifestyle medicine primary care physician had a patient covered by Medicare Advantage who was diagnosed with hyperlipidemia. The patient had total cholesterol of 226 and a triglycerides level of 132. Instead of prescribing the routine statin, the physician prescribed lifestyle behavior modifications. Within 3 weeks, the patient›s total cholesterol improved to 171 and triglycerides to 75. This was a great success for the delighted patient. However, the CMS 5-Star Rating System assigned the primary care physician a grade of C rather than A, which put the physician›s 5-star rating at risk. Why? Because the system bases its score largely on medication compliance. The physician was penalized despite achieving the optimal health outcome, and at a lower cost than with medication. This misalignment does not incentivize patient-centered care because it disregards patient preference, shared decision-making, and evidence-based practice.
Risk adjustment
Rather than automatically managing disease with ever-increasing quantities of costly medications and procedures, lifestyle medicine clinicians first pursue a goal of health restoration when appropriate. But Medicare risk adjustment incentivizes physicians to manage rather than reverse disease. How much Medicare pays health plans is determined in part by how sick the patients are; the sicker the patient, the more Medicare pays, because those patients› costs are expected to be higher. This ensures that health plans are not penalized for enrolling sicker patients. But a physician utilizing diet alone to achieve remission in a patient with type 2 diabetes is penalized financially because, when the risk is adjusted, diabetes is no longer listed among the patient›s conditions. So, Medicare pays the physician less money. That misalignment incentivizes clinicians to manage the symptoms of type 2 diabetes rather than achieve remission, despite remission being the ideal clinical outcome.
Realigning quality measures
Quality measures were developed to quantify health care processes and outcomes, and to ensure the delivery of safe care to all patients. However, over time the number of quality measures has swelled to 2500, evolving into a confusing, time-consuming, and even soul-crushing responsibility for the physician.
Instead of relying heavily on process measures, we must incentivize outcome measures that honor patient autonomy and allow clinicians to offer lifestyle intervention as the first line of treatment. Risk-score calculations should be adjusted so that we stop incentivizing disease management and penalizing disease reversal.
CMS’s proposed development of “a universal foundation” of quality measures is an opportunity to begin the realignment of quality measures and values. This foundation is intended to establish more consistent and meaningful measures, reduce clinician burnout by streamlining the reporting process, and advance health equity. For this change to be successful, it is vital that lifestyle behavior interventions – optimal nutrition, physical activity, restorative sleep, social connections, stress management, and avoidance of harmful substances – become the foundation of universal quality measures. This will ensure that every clinician is incentivized to discuss lifestyle behaviors with patients and pursue the first clinical step recommended by clinical practice guidelines for most chronic diseases. Only then can we truly deliver high-value, whole-person, person-centered care and achieve the quintuple aim.
Dr. Patel is president-elect, American College of Lifestyle Medicine; Lifestyle Medicine Medical Director, Wellvana Health, Midland, Tex. She has disclosed no relevant financial relationships.
A version of this article first appeared on Medscape.com.
The Centers for Medicare & Medicaid Services 2022 National Quality Strategy is described as an “ambitious long-term initiative that aims to promote the highest quality outcomes and safest care for all individuals.” It is a commendable goal for an overburdened U.S. health care system that spends more than other high-income counties yet experiences poorer outcomes. But whole-person, person-centered care cannot be achieved under current misaligned quality measures that fail to measure what we purport to value: the quintuple aim of improved health outcomes, cost savings, patient satisfaction, clinician well-being, and health equity.
Lifestyle first
Clinical practice guidelines for many chronic diseases recommend lifestyle intervention as the first and optimal treatment. A growing body of evidence supports lifestyle behavior interventions to treat and, when used intensively, even reverse common chronic conditions such as cardiovascular disease, obesity, and type 2 diabetes, while also providing effective prevention for those conditions. However, no current quality measures consider lifestyle interventions. In fact, some quality measures unintentionally penalize physicians for successfully treating or reversing disease through lifestyle behavior interventions while rewarding clinicians for meeting process measures – usually adherence to medication – regardless of whether health outcomes improved.
Rewarding medication adherence for the treatment of diseases in which lifestyle is a primary therapy (such as hypertension), combined with other health care constraints (lack of lifestyle education, time to spend with patients, and infrastructure support) incentivizes physicians to skip the conversation about lifestyle changes and go straight to medication prescription. Meanwhile, the clinician who takes the extra time to guide a patient toward lifestyle interventions that could treat their current disease and prevent future diseases – without side effects – is penalized.
Misaligned quality measures like these can stifle clinical judgment and risk reducing the practice of medicine to mindless box-checking. In many cases, patients are not even informed that lifestyle behavior change may be a treatment option (much less the first recommended option) for their conditions. This delivery of care is not person-centered and, in fact, may raise questions about the adequacy of informed treatment consent.
Reimbursement barriers
Lifestyle medicine is a growing medical specialty that uses therapeutic lifestyle interventions as a primary modality to treat chronic conditions. Since certification began in 2017, almost 2500 US physicians and 1000 nonphysician health professionals have earned certification. Health systems, including the U.S. military, are increasingly integrating lifestyle medicine. There have been advancements since one survey found that more than half of lifestyle medicine clinicians reported receiving no reimbursement for lifestyle behavior interventions. However, barriers, especially in fee-for-service systems, still inhibit many patients from receiving insurance coverage for comprehensive, interdisciplinary, and whole-person treatments called intensive therapeutic lifestyle change (ITLC) programs.
Existing comprehensive lifestyle programs that patients are eligible for (ie, the Diabetes Prevention Program and intensive behavioral therapy) are often so poorly reimbursed that clinicians and health systems decline to offer them. An example of a well-reimbursed ITLC program is intensive cardiac rehabilitation (ICR), which remains underutilized and limited to a narrow segment of patients, despite ICR›s proven benefits for managing comorbid risk factors such as hemoglobin A1c and weight. Even when lifestyle intervention programs are available and patients are eligible to participate (often through shared medical appointments), patient copays for the frequent visits required to achieve and sustain behavior change – or the lack of reimbursement for interdisciplinary team members – discourage engagement.
Penalizing successful outcomes
Despite the fact that lifestyle behaviors are top contributors to health and, conversely, contribute to up to 80% of chronic diseases, few quality measures focus on screening for lifestyle factors or treating diseases with lifestyle interventions. An example of an existing quality measure is screening or treatment for harmful substance use.
Specific quality measures that penalize lifestyle medicine approaches include pharmacotherapy for type 2 diabetes, dyslipidemia, osteoporosis, and gout as well as approaches to rheumatoid arthritis.
Statins offer a useful example of the conundrum faced by clinicians who want to offer lifestyle interventions. A lifestyle medicine primary care physician had a patient covered by Medicare Advantage who was diagnosed with hyperlipidemia. The patient had total cholesterol of 226 and a triglycerides level of 132. Instead of prescribing the routine statin, the physician prescribed lifestyle behavior modifications. Within 3 weeks, the patient›s total cholesterol improved to 171 and triglycerides to 75. This was a great success for the delighted patient. However, the CMS 5-Star Rating System assigned the primary care physician a grade of C rather than A, which put the physician›s 5-star rating at risk. Why? Because the system bases its score largely on medication compliance. The physician was penalized despite achieving the optimal health outcome, and at a lower cost than with medication. This misalignment does not incentivize patient-centered care because it disregards patient preference, shared decision-making, and evidence-based practice.
Risk adjustment
Rather than automatically managing disease with ever-increasing quantities of costly medications and procedures, lifestyle medicine clinicians first pursue a goal of health restoration when appropriate. But Medicare risk adjustment incentivizes physicians to manage rather than reverse disease. How much Medicare pays health plans is determined in part by how sick the patients are; the sicker the patient, the more Medicare pays, because those patients› costs are expected to be higher. This ensures that health plans are not penalized for enrolling sicker patients. But a physician utilizing diet alone to achieve remission in a patient with type 2 diabetes is penalized financially because, when the risk is adjusted, diabetes is no longer listed among the patient›s conditions. So, Medicare pays the physician less money. That misalignment incentivizes clinicians to manage the symptoms of type 2 diabetes rather than achieve remission, despite remission being the ideal clinical outcome.
Realigning quality measures
Quality measures were developed to quantify health care processes and outcomes, and to ensure the delivery of safe care to all patients. However, over time the number of quality measures has swelled to 2500, evolving into a confusing, time-consuming, and even soul-crushing responsibility for the physician.
Instead of relying heavily on process measures, we must incentivize outcome measures that honor patient autonomy and allow clinicians to offer lifestyle intervention as the first line of treatment. Risk-score calculations should be adjusted so that we stop incentivizing disease management and penalizing disease reversal.
CMS’s proposed development of “a universal foundation” of quality measures is an opportunity to begin the realignment of quality measures and values. This foundation is intended to establish more consistent and meaningful measures, reduce clinician burnout by streamlining the reporting process, and advance health equity. For this change to be successful, it is vital that lifestyle behavior interventions – optimal nutrition, physical activity, restorative sleep, social connections, stress management, and avoidance of harmful substances – become the foundation of universal quality measures. This will ensure that every clinician is incentivized to discuss lifestyle behaviors with patients and pursue the first clinical step recommended by clinical practice guidelines for most chronic diseases. Only then can we truly deliver high-value, whole-person, person-centered care and achieve the quintuple aim.
Dr. Patel is president-elect, American College of Lifestyle Medicine; Lifestyle Medicine Medical Director, Wellvana Health, Midland, Tex. She has disclosed no relevant financial relationships.
A version of this article first appeared on Medscape.com.