Commentary

The War on Cancer, declared by President Richard Nixon some 50 years ago, has been canceled during the second Trump administration in 2025 — so saith The New York Times Sunday magazine cover story on September 14, 2025. This war seems now to be best described as "The War on Cancer Research."

To our horror and disbelief, we've witnessed the slow but persistent drift of much of the United States citizenry away from science and the sudden and severe movement of the US government to crush much medical research. But it is not as if we were not warned.

In August 2024, on these pages and without political bias, I urged Medscape readers to pay attention to Project 2025. A great deal of what we as a population are now experiencing was laid out as a carefully constructed plan.

What is surprising is the cruel ruthlessness of the "move fast and break things" approach, taken with little apparent concern about the resultant human tragedies (workforce and patients) and no clear care about the resulting fallout. As we've now learned, destroying something as grand as our cancer research enterprise can be accomplished very quickly. Rebuilding it is certain to be slow and difficult and perhaps can never be accomplished.

In this new anti-science, anti-research, and anti-researcher reality, what can we now do?

First and foremost, we must recognize that the war on cancer is not over. Cancer is not canceled, even if much of the US government's research effort/funding has been. Those of us in medicine and public health often speak in quantification of causes of death of our populations. As such, I'll remind Medscape readers that cancer afflicts some 20 million humans worldwide each year, killing nearly 10 million. Although two-thirds of Americans diagnosed with a potentially lethal malignancy are cured, cancer still kills roughly 600,000 Americans each year. Cancer has been the second most frequent cause of death of Americans for 75 years.

Being inevitable and immutable, death itself is not the enemy. We all die. Disease, disability, pain, and human suffering are the real enemies of us all. Cancer maims, pains, diabetes, and torments some 20 million humans worldwide each year. That is a huge humanitarian problem that should be recognized by individuals of all creeds and backgrounds.

With this depletion of our domestic government basic and applied cancer research program, what can we do?

1. Think globally and look to the international scientific research enterprises — relying on them, much as they have relied on us.
2. Defend the universal importance of reliable and available literature on medical science.
3. Continue to translate and apply the vast amount of available published research in clinical practice and publish the results.
4. Urge private industries to expand their research budgets into areas of study that may not produce quickly tangible positive bottom-line results.
5. Remind the Secretary of the Department of Health and Human Services (for whom chronic diseases seem paramount) that cancer is the second leading American chronic disease by morbidity.
6. Redouble efforts of cancer prevention, especially urging the FDA to ban combustible tobacco and strive more diligently to decrease obesity.
7. Appeal to our vast philanthropic universe to increase its funding of nonprofit organizations active in the cancer investigation, diagnosis, and management space.

One such 501c3 organization is California-based Cancer Commons. (Disclosure: I named it in 2010 and serve as its editor in chief).

A commons is a space shared by a community to use for the common interest. As we originally envisioned it, a cancer commons is an open access internet location where individuals and organizations (eg, corporations, universities, government agencies, philanthropies) will voluntarily share their data to work together to defeat the common enemy of humans: cancer.

On September 8, 2025, Cancer Commons was the 15th annual Lundberg Institute Lecturer at the Commonwealth Club of California in San Francisco. At the lecture, Cancer Commons founder (and long-term survivor of metastatic malignant melanoma), Jay Martin "Martin" Tenenbaum, PhD, spoke of the need for a cancer commons and the founder's vision. Emma Shtivelman, PhD, the long-time compassionate chief scientist, described some of the thousands of patients with advanced cancer that she has helped — all free of charge. And newly named CEO Clifford Reid, MBA, PhD, used his entrepreneurial prowess to envision an ambitious future.

Cancer Commons has always focused on patients with cancer who are beyond standards of curative care. As Cancer Commons evolves, it anticipates focusing on patients with cancer who are beyond National Comprehensive Cancer Network Guidelines. The organization intends to greatly expand its 1000 patients per year with "high touch" engagement with PhD clinical scientists to many thousands by including artificial intelligence. It plans to extend its N-of-One approach to create new knowledge — especially regarding the hundreds of drugs that are FDA-approved for use in treating cancer but have not been further assessed for the utility in actually treating patients with cancer.

The war on cancer is not over. It remains a persistent foe that causes immense disability, pain, and human suffering. With government support depleted, the burden now shifts to the private sector and philanthropic organizations, such as Cancer Commons, to serve as the new vital infrastructure in the fight for a cure. Now, we must redouble our efforts to ensure that these research endeavors are supported if the US government will not do its part.

A version of this article first appeared on Medscape.com.

The War on Cancer, declared by President Richard Nixon some 50 years ago, has been canceled during the second Trump administration in 2025 — so saith The New York Times Sunday magazine cover story on September 14, 2025. This war seems now to be best described as "The War on Cancer Research."

To our horror and disbelief, we've witnessed the slow but persistent drift of much of the United States citizenry away from science and the sudden and severe movement of the US government to crush much medical research. But it is not as if we were not warned.

In August 2024, on these pages and without political bias, I urged Medscape readers to pay attention to Project 2025. A great deal of what we as a population are now experiencing was laid out as a carefully constructed plan.

What is surprising is the cruel ruthlessness of the "move fast and break things" approach, taken with little apparent concern about the resultant human tragedies (workforce and patients) and no clear care about the resulting fallout. As we've now learned, destroying something as grand as our cancer research enterprise can be accomplished very quickly. Rebuilding it is certain to be slow and difficult and perhaps can never be accomplished.

In this new anti-science, anti-research, and anti-researcher reality, what can we now do?

First and foremost, we must recognize that the war on cancer is not over. Cancer is not canceled, even if much of the US government's research effort/funding has been. Those of us in medicine and public health often speak in quantification of causes of death of our populations. As such, I'll remind Medscape readers that cancer afflicts some 20 million humans worldwide each year, killing nearly 10 million. Although two-thirds of Americans diagnosed with a potentially lethal malignancy are cured, cancer still kills roughly 600,000 Americans each year. Cancer has been the second most frequent cause of death of Americans for 75 years.

Being inevitable and immutable, death itself is not the enemy. We all die. Disease, disability, pain, and human suffering are the real enemies of us all. Cancer maims, pains, diabetes, and torments some 20 million humans worldwide each year. That is a huge humanitarian problem that should be recognized by individuals of all creeds and backgrounds.

With this depletion of our domestic government basic and applied cancer research program, what can we do?

1. Think globally and look to the international scientific research enterprises — relying on them, much as they have relied on us.
2. Defend the universal importance of reliable and available literature on medical science.
3. Continue to translate and apply the vast amount of available published research in clinical practice and publish the results.
4. Urge private industries to expand their research budgets into areas of study that may not produce quickly tangible positive bottom-line results.
5. Remind the Secretary of the Department of Health and Human Services (for whom chronic diseases seem paramount) that cancer is the second leading American chronic disease by morbidity.
6. Redouble efforts of cancer prevention, especially urging the FDA to ban combustible tobacco and strive more diligently to decrease obesity.
7. Appeal to our vast philanthropic universe to increase its funding of nonprofit organizations active in the cancer investigation, diagnosis, and management space.

One such 501c3 organization is California-based Cancer Commons. (Disclosure: I named it in 2010 and serve as its editor in chief).

A commons is a space shared by a community to use for the common interest. As we originally envisioned it, a cancer commons is an open access internet location where individuals and organizations (eg, corporations, universities, government agencies, philanthropies) will voluntarily share their data to work together to defeat the common enemy of humans: cancer.

On September 8, 2025, Cancer Commons was the 15th annual Lundberg Institute Lecturer at the Commonwealth Club of California in San Francisco. At the lecture, Cancer Commons founder (and long-term survivor of metastatic malignant melanoma), Jay Martin "Martin" Tenenbaum, PhD, spoke of the need for a cancer commons and the founder's vision. Emma Shtivelman, PhD, the long-time compassionate chief scientist, described some of the thousands of patients with advanced cancer that she has helped — all free of charge. And newly named CEO Clifford Reid, MBA, PhD, used his entrepreneurial prowess to envision an ambitious future.

Cancer Commons has always focused on patients with cancer who are beyond standards of curative care. As Cancer Commons evolves, it anticipates focusing on patients with cancer who are beyond National Comprehensive Cancer Network Guidelines. The organization intends to greatly expand its 1000 patients per year with "high touch" engagement with PhD clinical scientists to many thousands by including artificial intelligence. It plans to extend its N-of-One approach to create new knowledge — especially regarding the hundreds of drugs that are FDA-approved for use in treating cancer but have not been further assessed for the utility in actually treating patients with cancer.

The war on cancer is not over. It remains a persistent foe that causes immense disability, pain, and human suffering. With government support depleted, the burden now shifts to the private sector and philanthropic organizations, such as Cancer Commons, to serve as the new vital infrastructure in the fight for a cure. Now, we must redouble our efforts to ensure that these research endeavors are supported if the US government will not do its part.

A version of this article first appeared on Medscape.com.

The War on Cancer, declared by President Richard Nixon some 50 years ago, has been canceled during the second Trump administration in 2025 — so saith The New York Times Sunday magazine cover story on September 14, 2025. This war seems now to be best described as "The War on Cancer Research."

To our horror and disbelief, we've witnessed the slow but persistent drift of much of the United States citizenry away from science and the sudden and severe movement of the US government to crush much medical research. But it is not as if we were not warned.

In August 2024, on these pages and without political bias, I urged Medscape readers to pay attention to Project 2025. A great deal of what we as a population are now experiencing was laid out as a carefully constructed plan.

What is surprising is the cruel ruthlessness of the "move fast and break things" approach, taken with little apparent concern about the resultant human tragedies (workforce and patients) and no clear care about the resulting fallout. As we've now learned, destroying something as grand as our cancer research enterprise can be accomplished very quickly. Rebuilding it is certain to be slow and difficult and perhaps can never be accomplished.

In this new anti-science, anti-research, and anti-researcher reality, what can we now do?

First and foremost, we must recognize that the war on cancer is not over. Cancer is not canceled, even if much of the US government's research effort/funding has been. Those of us in medicine and public health often speak in quantification of causes of death of our populations. As such, I'll remind Medscape readers that cancer afflicts some 20 million humans worldwide each year, killing nearly 10 million. Although two-thirds of Americans diagnosed with a potentially lethal malignancy are cured, cancer still kills roughly 600,000 Americans each year. Cancer has been the second most frequent cause of death of Americans for 75 years.

Being inevitable and immutable, death itself is not the enemy. We all die. Disease, disability, pain, and human suffering are the real enemies of us all. Cancer maims, pains, diabetes, and torments some 20 million humans worldwide each year. That is a huge humanitarian problem that should be recognized by individuals of all creeds and backgrounds.

With this depletion of our domestic government basic and applied cancer research program, what can we do?

1. Think globally and look to the international scientific research enterprises — relying on them, much as they have relied on us.
2. Defend the universal importance of reliable and available literature on medical science.
3. Continue to translate and apply the vast amount of available published research in clinical practice and publish the results.
4. Urge private industries to expand their research budgets into areas of study that may not produce quickly tangible positive bottom-line results.
5. Remind the Secretary of the Department of Health and Human Services (for whom chronic diseases seem paramount) that cancer is the second leading American chronic disease by morbidity.
6. Redouble efforts of cancer prevention, especially urging the FDA to ban combustible tobacco and strive more diligently to decrease obesity.
7. Appeal to our vast philanthropic universe to increase its funding of nonprofit organizations active in the cancer investigation, diagnosis, and management space.

One such 501c3 organization is California-based Cancer Commons. (Disclosure: I named it in 2010 and serve as its editor in chief).

A commons is a space shared by a community to use for the common interest. As we originally envisioned it, a cancer commons is an open access internet location where individuals and organizations (eg, corporations, universities, government agencies, philanthropies) will voluntarily share their data to work together to defeat the common enemy of humans: cancer.

On September 8, 2025, Cancer Commons was the 15th annual Lundberg Institute Lecturer at the Commonwealth Club of California in San Francisco. At the lecture, Cancer Commons founder (and long-term survivor of metastatic malignant melanoma), Jay Martin "Martin" Tenenbaum, PhD, spoke of the need for a cancer commons and the founder's vision. Emma Shtivelman, PhD, the long-time compassionate chief scientist, described some of the thousands of patients with advanced cancer that she has helped — all free of charge. And newly named CEO Clifford Reid, MBA, PhD, used his entrepreneurial prowess to envision an ambitious future.

Cancer Commons has always focused on patients with cancer who are beyond standards of curative care. As Cancer Commons evolves, it anticipates focusing on patients with cancer who are beyond National Comprehensive Cancer Network Guidelines. The organization intends to greatly expand its 1000 patients per year with "high touch" engagement with PhD clinical scientists to many thousands by including artificial intelligence. It plans to extend its N-of-One approach to create new knowledge — especially regarding the hundreds of drugs that are FDA-approved for use in treating cancer but have not been further assessed for the utility in actually treating patients with cancer.

The war on cancer is not over. It remains a persistent foe that causes immense disability, pain, and human suffering. With government support depleted, the burden now shifts to the private sector and philanthropic organizations, such as Cancer Commons, to serve as the new vital infrastructure in the fight for a cure. Now, we must redouble our efforts to ensure that these research endeavors are supported if the US government will not do its part.

A version of this article first appeared on Medscape.com.

The practice of medicine is evolving rapidly, with clinicians facing enhanced pressure to maximize productivity while managing increasingly complex patients and related clinical documentation. Indeed, clinicians are spending less time seeing patients, and more time in front of a computer screen.

Despite the many rewards of clinical medicine, rates of clinical practice attrition have increased among physicians in all specialties since 2013 with enhanced administrative burdens identified as a prominent driver. Among its many applications, artificial intelligence (AI) has immense potential to reduce the administrative and cognitive burdens that contribute to clinician burnout and attrition through tools such as AI scribes – these technologies have been rapidly adopted across healthcare systems and are already in use by ~30% of physician practices. The hope is that AI scribes will significantly reduce documentation time, leading to improvements in clinician wellbeing and expanding capacity for patient care. Indeed, some studies have shown up to a 20-30% improvement in documentation efficiency. 

So, is AI a cure for physician burnout? The answer depends on what is done with these efficiency gains. If healthcare organizations respond to this enhanced efficiency by increasing patient volume expectations rather than allowing clinicians to recapture some of this time for meaningful work and professional wellbeing, it could create a so-called “workload paradox” where modest time savings are offset by greater productivity demands and the cognitive burden of reviewing AI-generated errors. While AI offers substantial promise as a tool to reduce administrative burdens, its success will depend on thoughtful and responsible implementation that prioritizes clinician well-being and patient safety in addition to productivity.

In our final issue of 2025, we highlight a recent RCT from Annals of Internal Medicine finding that fecal microbiota transplantation is at least as effective as vancomycin in treating primary C. difficile infection. In this month’s Member Spotlight, we feature Andrew Ofosu, MD, MPH (University of Cincinnati Health), who stresses the importance of transparency and compassion in communicating effectively with patients, particularly around complex diagnoses. We hope you enjoy this and all the exciting content in our December issue.

Megan A. Adams, MD, JD, MSc 

Editor in Chief

The practice of medicine is evolving rapidly, with clinicians facing enhanced pressure to maximize productivity while managing increasingly complex patients and related clinical documentation. Indeed, clinicians are spending less time seeing patients, and more time in front of a computer screen.

Despite the many rewards of clinical medicine, rates of clinical practice attrition have increased among physicians in all specialties since 2013 with enhanced administrative burdens identified as a prominent driver. Among its many applications, artificial intelligence (AI) has immense potential to reduce the administrative and cognitive burdens that contribute to clinician burnout and attrition through tools such as AI scribes – these technologies have been rapidly adopted across healthcare systems and are already in use by ~30% of physician practices. The hope is that AI scribes will significantly reduce documentation time, leading to improvements in clinician wellbeing and expanding capacity for patient care. Indeed, some studies have shown up to a 20-30% improvement in documentation efficiency. 

So, is AI a cure for physician burnout? The answer depends on what is done with these efficiency gains. If healthcare organizations respond to this enhanced efficiency by increasing patient volume expectations rather than allowing clinicians to recapture some of this time for meaningful work and professional wellbeing, it could create a so-called “workload paradox” where modest time savings are offset by greater productivity demands and the cognitive burden of reviewing AI-generated errors. While AI offers substantial promise as a tool to reduce administrative burdens, its success will depend on thoughtful and responsible implementation that prioritizes clinician well-being and patient safety in addition to productivity.

In our final issue of 2025, we highlight a recent RCT from Annals of Internal Medicine finding that fecal microbiota transplantation is at least as effective as vancomycin in treating primary C. difficile infection. In this month’s Member Spotlight, we feature Andrew Ofosu, MD, MPH (University of Cincinnati Health), who stresses the importance of transparency and compassion in communicating effectively with patients, particularly around complex diagnoses. We hope you enjoy this and all the exciting content in our December issue.

Megan A. Adams, MD, JD, MSc 

Editor in Chief

The practice of medicine is evolving rapidly, with clinicians facing enhanced pressure to maximize productivity while managing increasingly complex patients and related clinical documentation. Indeed, clinicians are spending less time seeing patients, and more time in front of a computer screen.

Despite the many rewards of clinical medicine, rates of clinical practice attrition have increased among physicians in all specialties since 2013 with enhanced administrative burdens identified as a prominent driver. Among its many applications, artificial intelligence (AI) has immense potential to reduce the administrative and cognitive burdens that contribute to clinician burnout and attrition through tools such as AI scribes – these technologies have been rapidly adopted across healthcare systems and are already in use by ~30% of physician practices. The hope is that AI scribes will significantly reduce documentation time, leading to improvements in clinician wellbeing and expanding capacity for patient care. Indeed, some studies have shown up to a 20-30% improvement in documentation efficiency. 

So, is AI a cure for physician burnout? The answer depends on what is done with these efficiency gains. If healthcare organizations respond to this enhanced efficiency by increasing patient volume expectations rather than allowing clinicians to recapture some of this time for meaningful work and professional wellbeing, it could create a so-called “workload paradox” where modest time savings are offset by greater productivity demands and the cognitive burden of reviewing AI-generated errors. While AI offers substantial promise as a tool to reduce administrative burdens, its success will depend on thoughtful and responsible implementation that prioritizes clinician well-being and patient safety in addition to productivity.

In our final issue of 2025, we highlight a recent RCT from Annals of Internal Medicine finding that fecal microbiota transplantation is at least as effective as vancomycin in treating primary C. difficile infection. In this month’s Member Spotlight, we feature Andrew Ofosu, MD, MPH (University of Cincinnati Health), who stresses the importance of transparency and compassion in communicating effectively with patients, particularly around complex diagnoses. We hope you enjoy this and all the exciting content in our December issue.

Megan A. Adams, MD, JD, MSc 

Editor in Chief

Are you seeing any increase or trends in cutaneous adverse effects related to the use of GLP-1 agonists in your practice?

DR. SHIELDS: The use of GLP-1 agonists is increasing substantially across numerous populations. Patients are using these medications not only for weight management and diabetes control but also for blood pressure modulation and cardiovascular risk reduction. The market size is expected to grow at a rate of about 6% until 2027. While severe cutaneous adverse effects still are considered relatively rare with GLP-1 agonist use, mild adverse effects are quite common. Dermatologists should be familiar with these effects and how to manage them. Rare but serious cutaneous reactions include morbilliform drug eruptions, dermal hypersensitivity reactions, panniculitis, and bullous pemphigoid. It is thought that some GLP-1 agonists may cause more skin reactions than others; for example, exenatide extended-release has been associated with cutaneous adverse events more frequently than other GLP-1 agonists in a recent comprehensive literature review.

Do you see a role for dermatologists in monitoring or managing the downstream dermatologic effects of GLP-1 agonists over the next few years?

DR. SHIELDS: Absolutely. When patients develop a drug eruption, bullous pemphigoid, or eosinophilic panniculitis, dermatologists are going to be the ones to diagnose and manage therapy. Awareness of these adverse effects is crucial to timely and thoughtful discussions surrounding medication discontinuation vs a “treat through” approach.

Do you recommend coordinating with endocrinologists or obesity medicine specialists when managing shared patients on GLP-1s (particularly if skin concerns arise)?

DR. SHIELDS: Yes. This is crucial to patient success. Co-management can provide clarity around the indication for therapy and allow for a thoughtful risk-benefit discussion with the patient, primary care physician, endocrinologist, cardiologist, etc. In my practice, I have found that many patients do not want to stop therapy even when they develop cutaneous adverse effects. There are options to transition therapy or treat through in some cases, but having a comprehensive monitoring and therapy plan is critical.

Have you encountered cases in which rapid weight loss from GLP-1s worsened conditions such as loose skin, cellulite, or facial lipoatrophy, leading to new aesthetic concerns? How would you recommend counseling and/or treating affected patients?

DR. SHIELDS: Accelerated facial aging is a noticeable adverse effect in patients who undergo treatment with GLP-1 agonists, especially when used off-label for weight loss. Localized loss of facial fat can result in altered facial proportions and excess skin. There are multiple additional mechanisms that may underlie accelerated facial aging in patients on GLP-1s, and really we are just beginning to scratch the surface of why and how this happens. Understanding these mechanisms will open the door to downstream preventive and therapeutic options. If patients experience new aesthetic concerns, I currently work with them to adjust their medication to slow weight loss, recommend improved nutrition and hydration, encourage exercise and weight training to maintain muscle mass, and engage my cosmetic dermatology colleagues to discuss procedures such as dermal fillers.

All patients starting GLP-1 agonists should be thoroughly counseled on risks and adverse effects of their medication. These are well reported and should be considered carefully. Starting with lower medication dosing in conjunction with slow escalation and careful monitoring can be helpful in combatting these adverse effects.

Are you seeing any increase or trends in cutaneous adverse effects related to the use of GLP-1 agonists in your practice?

DR. SHIELDS: The use of GLP-1 agonists is increasing substantially across numerous populations. Patients are using these medications not only for weight management and diabetes control but also for blood pressure modulation and cardiovascular risk reduction. The market size is expected to grow at a rate of about 6% until 2027. While severe cutaneous adverse effects still are considered relatively rare with GLP-1 agonist use, mild adverse effects are quite common. Dermatologists should be familiar with these effects and how to manage them. Rare but serious cutaneous reactions include morbilliform drug eruptions, dermal hypersensitivity reactions, panniculitis, and bullous pemphigoid. It is thought that some GLP-1 agonists may cause more skin reactions than others; for example, exenatide extended-release has been associated with cutaneous adverse events more frequently than other GLP-1 agonists in a recent comprehensive literature review.

Do you see a role for dermatologists in monitoring or managing the downstream dermatologic effects of GLP-1 agonists over the next few years?

DR. SHIELDS: Absolutely. When patients develop a drug eruption, bullous pemphigoid, or eosinophilic panniculitis, dermatologists are going to be the ones to diagnose and manage therapy. Awareness of these adverse effects is crucial to timely and thoughtful discussions surrounding medication discontinuation vs a “treat through” approach.

Do you recommend coordinating with endocrinologists or obesity medicine specialists when managing shared patients on GLP-1s (particularly if skin concerns arise)?

DR. SHIELDS: Yes. This is crucial to patient success. Co-management can provide clarity around the indication for therapy and allow for a thoughtful risk-benefit discussion with the patient, primary care physician, endocrinologist, cardiologist, etc. In my practice, I have found that many patients do not want to stop therapy even when they develop cutaneous adverse effects. There are options to transition therapy or treat through in some cases, but having a comprehensive monitoring and therapy plan is critical.

Have you encountered cases in which rapid weight loss from GLP-1s worsened conditions such as loose skin, cellulite, or facial lipoatrophy, leading to new aesthetic concerns? How would you recommend counseling and/or treating affected patients?

DR. SHIELDS: Accelerated facial aging is a noticeable adverse effect in patients who undergo treatment with GLP-1 agonists, especially when used off-label for weight loss. Localized loss of facial fat can result in altered facial proportions and excess skin. There are multiple additional mechanisms that may underlie accelerated facial aging in patients on GLP-1s, and really we are just beginning to scratch the surface of why and how this happens. Understanding these mechanisms will open the door to downstream preventive and therapeutic options. If patients experience new aesthetic concerns, I currently work with them to adjust their medication to slow weight loss, recommend improved nutrition and hydration, encourage exercise and weight training to maintain muscle mass, and engage my cosmetic dermatology colleagues to discuss procedures such as dermal fillers.

All patients starting GLP-1 agonists should be thoroughly counseled on risks and adverse effects of their medication. These are well reported and should be considered carefully. Starting with lower medication dosing in conjunction with slow escalation and careful monitoring can be helpful in combatting these adverse effects.

Are you seeing any increase or trends in cutaneous adverse effects related to the use of GLP-1 agonists in your practice?

DR. SHIELDS: The use of GLP-1 agonists is increasing substantially across numerous populations. Patients are using these medications not only for weight management and diabetes control but also for blood pressure modulation and cardiovascular risk reduction. The market size is expected to grow at a rate of about 6% until 2027. While severe cutaneous adverse effects still are considered relatively rare with GLP-1 agonist use, mild adverse effects are quite common. Dermatologists should be familiar with these effects and how to manage them. Rare but serious cutaneous reactions include morbilliform drug eruptions, dermal hypersensitivity reactions, panniculitis, and bullous pemphigoid. It is thought that some GLP-1 agonists may cause more skin reactions than others; for example, exenatide extended-release has been associated with cutaneous adverse events more frequently than other GLP-1 agonists in a recent comprehensive literature review.

Do you see a role for dermatologists in monitoring or managing the downstream dermatologic effects of GLP-1 agonists over the next few years?

DR. SHIELDS: Absolutely. When patients develop a drug eruption, bullous pemphigoid, or eosinophilic panniculitis, dermatologists are going to be the ones to diagnose and manage therapy. Awareness of these adverse effects is crucial to timely and thoughtful discussions surrounding medication discontinuation vs a “treat through” approach.

Do you recommend coordinating with endocrinologists or obesity medicine specialists when managing shared patients on GLP-1s (particularly if skin concerns arise)?

DR. SHIELDS: Yes. This is crucial to patient success. Co-management can provide clarity around the indication for therapy and allow for a thoughtful risk-benefit discussion with the patient, primary care physician, endocrinologist, cardiologist, etc. In my practice, I have found that many patients do not want to stop therapy even when they develop cutaneous adverse effects. There are options to transition therapy or treat through in some cases, but having a comprehensive monitoring and therapy plan is critical.

Have you encountered cases in which rapid weight loss from GLP-1s worsened conditions such as loose skin, cellulite, or facial lipoatrophy, leading to new aesthetic concerns? How would you recommend counseling and/or treating affected patients?

DR. SHIELDS: Accelerated facial aging is a noticeable adverse effect in patients who undergo treatment with GLP-1 agonists, especially when used off-label for weight loss. Localized loss of facial fat can result in altered facial proportions and excess skin. There are multiple additional mechanisms that may underlie accelerated facial aging in patients on GLP-1s, and really we are just beginning to scratch the surface of why and how this happens. Understanding these mechanisms will open the door to downstream preventive and therapeutic options. If patients experience new aesthetic concerns, I currently work with them to adjust their medication to slow weight loss, recommend improved nutrition and hydration, encourage exercise and weight training to maintain muscle mass, and engage my cosmetic dermatology colleagues to discuss procedures such as dermal fillers.

All patients starting GLP-1 agonists should be thoroughly counseled on risks and adverse effects of their medication. These are well reported and should be considered carefully. Starting with lower medication dosing in conjunction with slow escalation and careful monitoring can be helpful in combatting these adverse effects.

Atopic dermatitis (AD) is a chronic inflammatory skin condition associated with skin barrier impairment and immune system dysregulation.¹ Development of AD in young children can present challenges in determining appropriate treatment regimens. Natural remedies for AD often are promoted on social media over traditional treatments, including topical corticosteroids (TCSs), which can contribute to corticophobia.² Dermatologists play a critical role not only in optimizing topical therapy but also addressing patient interest in natural approaches to AD, including diet-related questions. This article outlines the role of diet and probiotics in pediatric AD and reviews the topical treatments currently approved for this patient population.

Diet and Probiotics

With a growing focus on natural therapies for AD, dietary interventions have come to the forefront. A prevalent theme among patients and their families is addressing gut health and allergic triggers. Broad elimination diets have not shown clinical benefit in patients with AD regardless of age,³ and in children, they may result in nutritional deficiencies, poor growth, and increased risk for IgE-mediated food allergies.⁴ If a true food allergy is identified based on positive IgE and an acute clinical reaction, elimination of the allergen may provide some benefit.⁵

The link between gut microbiota and skin health has driven an interest in the role of probiotics in the treatment of pediatric AD. A meta-analysis of 20 articles concluded that, whether administered to infants or breastfeeding mothers, use of probiotics overall led to a significant reduction in AD risk in infants (P=.001). Lactobacillus and mixed strains were effective.⁶ While broad elimination diets are not used to treat AD, probiotic supplementation can be considered for prevention of AD.

Topical Corticosteroids

Topical corticosteroids are the cornerstone of AD treatment; however, corticophobia among patients is on the rise, leading to poor adherence and suboptimal control of AD.⁷ Mild cutaneous adverse effects (AEs) including skin atrophy, striae, and telangiectasias may occur. Rarely, systemic AEs occur due to absorption of TCSs into the bloodstream, mainly with application of potent steroids over large body surface areas or under occlusion.⁸ When the optimal potency of a TCS is chosen and used appropriately, incidence of AEs from TCS use is very low.⁹

Counseling parents about risk factors that can lead to AEs during treatment with TCSs and formulating regimens that minimize these risks while maintaining efficacy increases adherence and outcomes. Pulse maintenance dosing of TCSs typically involves application 1 to 2 times weekly to areas of the skin that are prone to frequent outbreaks. Pulse maintenance dosing can reduce the incidence of AD flares while also decreasing the total amount of topical medication needed as compared to the reactive approach alone, thereby reducing risk for AEs.⁸

Steroid-Sparing Topical Treatments

Although TCSs are considered first-line agents, recently there has been an advent of steroid-sparing topical agents approved by the US Food and Drug Administration (FDA) for pediatric patients with AD, including topical calcineurin inhibitors (TCIs), phosphodiesterase 4 inhibitors, a Janus kinase inhibitor, and aryl hydrocarbon receptor agonists. Offering steroid-sparing agents in these patients can help ease parental anxiety regarding TCS overuse.

Topical Calcineurin Inhibitors—Pimecrolimus cream 1% and tacrolimus ointment 0.03% are approved for patients aged 2 years and older and have anti-inflammatory and antipruritic effects equivalent to low-potency TCS. Tacrolimus ­ointment 0.1% is approved for patients aged 16 years and older with similar efficacy to a midpotency TCSs. Pimecrolimus cream 1% and tacrolimus ointment 0.03% often are used off-label in ­children younger than 2 years, as supported by clinical trials showing their safety and efficacy.¹⁰ 

Topical calcineurin inhibitors can replace or supplement TCSs, making TCIs a desirable option for avoidance of steroid-related AEs. The addition of a TCI to spot treatment or a pulse regimen in a young patient can reassure them and their caregivers that the provider is proactively reducing the risk of TCS overuse. The largest barrier to TCI use is the FDA’s black box warning based on the oral formulation of tacrolimus, citing a potential increased risk for lymphoma and skin cancer; however, there is no evidence for substantial systemic absorption of topical pimecrolimus or tacrolimus.¹¹ Large task-force reviews have found no association between TCI use and development of malignancy.^12,13 Based on the current data, counseling patients and their caregivers that this risk primarily is theoretical may help them more confidently integrate TCIs into their treatment regimen. Burning and tingling may occur in a minority of pediatric patients using TCIs for AD. Applying the medication to open wounds or inflamed skin increases the risk for stinging, but pretreatment with a short course of TCSs before transitioning to a TCI may boost tolerance.¹⁴ 

Phosphodiesterase 4 Inhibitors—Crisaborole ointment 2%, a phosphodiesterase 4 inhibitor, is approved for children aged 3 months and older with mild to moderate AD. Its use has been more limited than TCSs and TCIs, as local irritation including stinging and burning can occur in up to 50% of patients.¹⁵ One study comparing crisaborole 2% with tacrolimus 0.03% revealed greater improvement with tacrolimus.¹⁶ A second phosphodiesterase 4 inhibitor approved for once-daily use in children aged 6 years and older with mild to moderate AD is roflumilast cream 0.15%. Roflumilast reduces eczema severity and pruritus, with AEs also limited to application-site stinging and burning.¹⁷ 

Janus Kinase Inhibitor—Ruxolitinib cream 1.5%, a Janus kinase inhibitor, has been approved by the FDA since 2023 for twice-daily use in children aged 12 years and older with AD. Similar to TCIs, ruxolitinib cream carries a black box warning. Short-term safety data on ruxolitinib cream have revealed low levels of ruxolitinib concentration in plasma¹⁸; however, long-term studies on topical Janus kinase inhibitors for AD in pediatric and adult populations are lacking. To reduce the risk for systemic absorption, recommendations include limiting usage to 60 g per week and limiting treatment to less than 20% of the body surface area.¹⁹ Ruxolitinib has efficacy similar to or possibly superior to triamcinolone 0.1%.²⁰ Ruxolitinib is emerging as a promising nonsteroidal option that potentially is highly efficacious and well tolerated without cutaneous AEs.  

Aryl Hydrocarbon Receptor Agonist—Tapinarof cream 1% is an aryl hydrocarbon receptor agonist that has been approved by the FDA since 2024 for children aged 2 years and older as a once-daily treatment for moderate to severe AD. Adverse events include folliculitis, nasopharyngitis, and headache, which are mostly mild or moderate.²¹

Final Thoughts

Topical management of pediatric AD includes traditional therapy with TCSs and newer steroid-sparing agents, which can help address corticophobia. Anticipatory guidance regarding the safety and long-term effects of individual therapies is critical to ensuring patient adherence to treatment regimens. Probiotics may help prevent pediatric AD, but future studies are needed to determine their role in treatment.

Atopic dermatitis (AD) is a chronic inflammatory skin condition associated with skin barrier impairment and immune system dysregulation.¹ Development of AD in young children can present challenges in determining appropriate treatment regimens. Natural remedies for AD often are promoted on social media over traditional treatments, including topical corticosteroids (TCSs), which can contribute to corticophobia.² Dermatologists play a critical role not only in optimizing topical therapy but also addressing patient interest in natural approaches to AD, including diet-related questions. This article outlines the role of diet and probiotics in pediatric AD and reviews the topical treatments currently approved for this patient population.

Diet and Probiotics

With a growing focus on natural therapies for AD, dietary interventions have come to the forefront. A prevalent theme among patients and their families is addressing gut health and allergic triggers. Broad elimination diets have not shown clinical benefit in patients with AD regardless of age,³ and in children, they may result in nutritional deficiencies, poor growth, and increased risk for IgE-mediated food allergies.⁴ If a true food allergy is identified based on positive IgE and an acute clinical reaction, elimination of the allergen may provide some benefit.⁵

The link between gut microbiota and skin health has driven an interest in the role of probiotics in the treatment of pediatric AD. A meta-analysis of 20 articles concluded that, whether administered to infants or breastfeeding mothers, use of probiotics overall led to a significant reduction in AD risk in infants (P=.001). Lactobacillus and mixed strains were effective.⁶ While broad elimination diets are not used to treat AD, probiotic supplementation can be considered for prevention of AD.

Topical Corticosteroids

Topical corticosteroids are the cornerstone of AD treatment; however, corticophobia among patients is on the rise, leading to poor adherence and suboptimal control of AD.⁷ Mild cutaneous adverse effects (AEs) including skin atrophy, striae, and telangiectasias may occur. Rarely, systemic AEs occur due to absorption of TCSs into the bloodstream, mainly with application of potent steroids over large body surface areas or under occlusion.⁸ When the optimal potency of a TCS is chosen and used appropriately, incidence of AEs from TCS use is very low.⁹

Counseling parents about risk factors that can lead to AEs during treatment with TCSs and formulating regimens that minimize these risks while maintaining efficacy increases adherence and outcomes. Pulse maintenance dosing of TCSs typically involves application 1 to 2 times weekly to areas of the skin that are prone to frequent outbreaks. Pulse maintenance dosing can reduce the incidence of AD flares while also decreasing the total amount of topical medication needed as compared to the reactive approach alone, thereby reducing risk for AEs.⁸

Steroid-Sparing Topical Treatments

Although TCSs are considered first-line agents, recently there has been an advent of steroid-sparing topical agents approved by the US Food and Drug Administration (FDA) for pediatric patients with AD, including topical calcineurin inhibitors (TCIs), phosphodiesterase 4 inhibitors, a Janus kinase inhibitor, and aryl hydrocarbon receptor agonists. Offering steroid-sparing agents in these patients can help ease parental anxiety regarding TCS overuse.

Topical Calcineurin Inhibitors—Pimecrolimus cream 1% and tacrolimus ointment 0.03% are approved for patients aged 2 years and older and have anti-inflammatory and antipruritic effects equivalent to low-potency TCS. Tacrolimus ­ointment 0.1% is approved for patients aged 16 years and older with similar efficacy to a midpotency TCSs. Pimecrolimus cream 1% and tacrolimus ointment 0.03% often are used off-label in ­children younger than 2 years, as supported by clinical trials showing their safety and efficacy.¹⁰ 

Topical calcineurin inhibitors can replace or supplement TCSs, making TCIs a desirable option for avoidance of steroid-related AEs. The addition of a TCI to spot treatment or a pulse regimen in a young patient can reassure them and their caregivers that the provider is proactively reducing the risk of TCS overuse. The largest barrier to TCI use is the FDA’s black box warning based on the oral formulation of tacrolimus, citing a potential increased risk for lymphoma and skin cancer; however, there is no evidence for substantial systemic absorption of topical pimecrolimus or tacrolimus.¹¹ Large task-force reviews have found no association between TCI use and development of malignancy.^12,13 Based on the current data, counseling patients and their caregivers that this risk primarily is theoretical may help them more confidently integrate TCIs into their treatment regimen. Burning and tingling may occur in a minority of pediatric patients using TCIs for AD. Applying the medication to open wounds or inflamed skin increases the risk for stinging, but pretreatment with a short course of TCSs before transitioning to a TCI may boost tolerance.¹⁴ 

Phosphodiesterase 4 Inhibitors—Crisaborole ointment 2%, a phosphodiesterase 4 inhibitor, is approved for children aged 3 months and older with mild to moderate AD. Its use has been more limited than TCSs and TCIs, as local irritation including stinging and burning can occur in up to 50% of patients.¹⁵ One study comparing crisaborole 2% with tacrolimus 0.03% revealed greater improvement with tacrolimus.¹⁶ A second phosphodiesterase 4 inhibitor approved for once-daily use in children aged 6 years and older with mild to moderate AD is roflumilast cream 0.15%. Roflumilast reduces eczema severity and pruritus, with AEs also limited to application-site stinging and burning.¹⁷ 

Janus Kinase Inhibitor—Ruxolitinib cream 1.5%, a Janus kinase inhibitor, has been approved by the FDA since 2023 for twice-daily use in children aged 12 years and older with AD. Similar to TCIs, ruxolitinib cream carries a black box warning. Short-term safety data on ruxolitinib cream have revealed low levels of ruxolitinib concentration in plasma¹⁸; however, long-term studies on topical Janus kinase inhibitors for AD in pediatric and adult populations are lacking. To reduce the risk for systemic absorption, recommendations include limiting usage to 60 g per week and limiting treatment to less than 20% of the body surface area.¹⁹ Ruxolitinib has efficacy similar to or possibly superior to triamcinolone 0.1%.²⁰ Ruxolitinib is emerging as a promising nonsteroidal option that potentially is highly efficacious and well tolerated without cutaneous AEs.  

Aryl Hydrocarbon Receptor Agonist—Tapinarof cream 1% is an aryl hydrocarbon receptor agonist that has been approved by the FDA since 2024 for children aged 2 years and older as a once-daily treatment for moderate to severe AD. Adverse events include folliculitis, nasopharyngitis, and headache, which are mostly mild or moderate.²¹

Final Thoughts

Topical management of pediatric AD includes traditional therapy with TCSs and newer steroid-sparing agents, which can help address corticophobia. Anticipatory guidance regarding the safety and long-term effects of individual therapies is critical to ensuring patient adherence to treatment regimens. Probiotics may help prevent pediatric AD, but future studies are needed to determine their role in treatment.

Atopic dermatitis (AD) is a chronic inflammatory skin condition associated with skin barrier impairment and immune system dysregulation.¹ Development of AD in young children can present challenges in determining appropriate treatment regimens. Natural remedies for AD often are promoted on social media over traditional treatments, including topical corticosteroids (TCSs), which can contribute to corticophobia.² Dermatologists play a critical role not only in optimizing topical therapy but also addressing patient interest in natural approaches to AD, including diet-related questions. This article outlines the role of diet and probiotics in pediatric AD and reviews the topical treatments currently approved for this patient population.

Diet and Probiotics

With a growing focus on natural therapies for AD, dietary interventions have come to the forefront. A prevalent theme among patients and their families is addressing gut health and allergic triggers. Broad elimination diets have not shown clinical benefit in patients with AD regardless of age,³ and in children, they may result in nutritional deficiencies, poor growth, and increased risk for IgE-mediated food allergies.⁴ If a true food allergy is identified based on positive IgE and an acute clinical reaction, elimination of the allergen may provide some benefit.⁵

The link between gut microbiota and skin health has driven an interest in the role of probiotics in the treatment of pediatric AD. A meta-analysis of 20 articles concluded that, whether administered to infants or breastfeeding mothers, use of probiotics overall led to a significant reduction in AD risk in infants (P=.001). Lactobacillus and mixed strains were effective.⁶ While broad elimination diets are not used to treat AD, probiotic supplementation can be considered for prevention of AD.

Topical Corticosteroids

Topical corticosteroids are the cornerstone of AD treatment; however, corticophobia among patients is on the rise, leading to poor adherence and suboptimal control of AD.⁷ Mild cutaneous adverse effects (AEs) including skin atrophy, striae, and telangiectasias may occur. Rarely, systemic AEs occur due to absorption of TCSs into the bloodstream, mainly with application of potent steroids over large body surface areas or under occlusion.⁸ When the optimal potency of a TCS is chosen and used appropriately, incidence of AEs from TCS use is very low.⁹

Counseling parents about risk factors that can lead to AEs during treatment with TCSs and formulating regimens that minimize these risks while maintaining efficacy increases adherence and outcomes. Pulse maintenance dosing of TCSs typically involves application 1 to 2 times weekly to areas of the skin that are prone to frequent outbreaks. Pulse maintenance dosing can reduce the incidence of AD flares while also decreasing the total amount of topical medication needed as compared to the reactive approach alone, thereby reducing risk for AEs.⁸

Steroid-Sparing Topical Treatments

Although TCSs are considered first-line agents, recently there has been an advent of steroid-sparing topical agents approved by the US Food and Drug Administration (FDA) for pediatric patients with AD, including topical calcineurin inhibitors (TCIs), phosphodiesterase 4 inhibitors, a Janus kinase inhibitor, and aryl hydrocarbon receptor agonists. Offering steroid-sparing agents in these patients can help ease parental anxiety regarding TCS overuse.

Topical Calcineurin Inhibitors—Pimecrolimus cream 1% and tacrolimus ointment 0.03% are approved for patients aged 2 years and older and have anti-inflammatory and antipruritic effects equivalent to low-potency TCS. Tacrolimus ­ointment 0.1% is approved for patients aged 16 years and older with similar efficacy to a midpotency TCSs. Pimecrolimus cream 1% and tacrolimus ointment 0.03% often are used off-label in ­children younger than 2 years, as supported by clinical trials showing their safety and efficacy.¹⁰ 

Topical calcineurin inhibitors can replace or supplement TCSs, making TCIs a desirable option for avoidance of steroid-related AEs. The addition of a TCI to spot treatment or a pulse regimen in a young patient can reassure them and their caregivers that the provider is proactively reducing the risk of TCS overuse. The largest barrier to TCI use is the FDA’s black box warning based on the oral formulation of tacrolimus, citing a potential increased risk for lymphoma and skin cancer; however, there is no evidence for substantial systemic absorption of topical pimecrolimus or tacrolimus.¹¹ Large task-force reviews have found no association between TCI use and development of malignancy.^12,13 Based on the current data, counseling patients and their caregivers that this risk primarily is theoretical may help them more confidently integrate TCIs into their treatment regimen. Burning and tingling may occur in a minority of pediatric patients using TCIs for AD. Applying the medication to open wounds or inflamed skin increases the risk for stinging, but pretreatment with a short course of TCSs before transitioning to a TCI may boost tolerance.¹⁴ 

Phosphodiesterase 4 Inhibitors—Crisaborole ointment 2%, a phosphodiesterase 4 inhibitor, is approved for children aged 3 months and older with mild to moderate AD. Its use has been more limited than TCSs and TCIs, as local irritation including stinging and burning can occur in up to 50% of patients.¹⁵ One study comparing crisaborole 2% with tacrolimus 0.03% revealed greater improvement with tacrolimus.¹⁶ A second phosphodiesterase 4 inhibitor approved for once-daily use in children aged 6 years and older with mild to moderate AD is roflumilast cream 0.15%. Roflumilast reduces eczema severity and pruritus, with AEs also limited to application-site stinging and burning.¹⁷ 

Janus Kinase Inhibitor—Ruxolitinib cream 1.5%, a Janus kinase inhibitor, has been approved by the FDA since 2023 for twice-daily use in children aged 12 years and older with AD. Similar to TCIs, ruxolitinib cream carries a black box warning. Short-term safety data on ruxolitinib cream have revealed low levels of ruxolitinib concentration in plasma¹⁸; however, long-term studies on topical Janus kinase inhibitors for AD in pediatric and adult populations are lacking. To reduce the risk for systemic absorption, recommendations include limiting usage to 60 g per week and limiting treatment to less than 20% of the body surface area.¹⁹ Ruxolitinib has efficacy similar to or possibly superior to triamcinolone 0.1%.²⁰ Ruxolitinib is emerging as a promising nonsteroidal option that potentially is highly efficacious and well tolerated without cutaneous AEs.  

Aryl Hydrocarbon Receptor Agonist—Tapinarof cream 1% is an aryl hydrocarbon receptor agonist that has been approved by the FDA since 2024 for children aged 2 years and older as a once-daily treatment for moderate to severe AD. Adverse events include folliculitis, nasopharyngitis, and headache, which are mostly mild or moderate.²¹

Final Thoughts

Topical management of pediatric AD includes traditional therapy with TCSs and newer steroid-sparing agents, which can help address corticophobia. Anticipatory guidance regarding the safety and long-term effects of individual therapies is critical to ensuring patient adherence to treatment regimens. Probiotics may help prevent pediatric AD, but future studies are needed to determine their role in treatment.

Hi, everyone. I’m Dr Kenny Lin. I am a family physician and associate director of the Lancaster General Hospital Family Medicine Residency, and I blog at Common Sense Family Doctor.

The receding of the pandemic and the understandable desire to return to normalcy has made COVID-19 vaccines a lower priority for many of our patients. However, family physicians should keep in mind that from October 1, 2024, to September 6, 2025, COVID-19 was responsible for an estimated 3.2 to 4.6 million outpatient visits, 360,000 to 520,000 hospitalizations, and 42,000 to 60,000 deaths.

In a previous commentary, I discussed the worsening disconnect between the evidence supporting the effectiveness and safety of vaccinations and increasing reluctance of patients and parents to receive them, fueled by misinformation from federal health agencies and the packing of the Advisory Committee on Immunization Practices (ACIP) with vaccine skeptics. Since then, Secretary of Health and Human Services (HHS), Robert F. Kennedy, Jr, has fired Dr Susan Monarez, his handpicked director of the CDC. This caused three senior CDC officials to resign in protest and precipitated further turmoil at the embattled agency. 

The FDA has approved 3 updated COVID-19 vaccines targeted to currently circulating strains: an mRNA vaccine from Moderna (Spikevax) for those aged 6 months or older; an mRNA vaccine from Pfizer/BioNTech (Comirnaty) for those aged ≥ 5 years; and a protein subunit vaccine from Novavax (Nuvaxovid) for those aged ≥ 12 years. However, approvals restricting the scope of these approvals to certain high-risk groups, combined with the ACIP’s recent decision to not explicitly recommend them for any group, have complicated access for many patients.

Medical groups, including the American Academy of Pediatrics (AAP), the American Academy of Family Physicians (AAFP), and the American College of Obstetricians and Gynecologists (ACOG), have published their own recommendations (Table). Of note, in opposition to the FDA and ACIP, the AAP and AAFP strongly recommend routine vaccination for children aged 6 to 23 months because they have the highest risk for hospitalization. The AAFP and ACOG both recommend COVID-19 vaccination in pregnancy to protect the pregnant patient and provide passive antibody protection to their infants up to 6 months of age. The Vaccine Integrity Project’s review of 12 safety studies published since June 2024 found that mRNA vaccines were not associated with increases in any adverse maternal or infant outcomes and had a possible protective effect against preterm birth.

In my previous commentary, 70% of Medscape readers indicated that they would follow vaccination recommendations from AAP even if they differed from CDC guidance. Administering vaccines outside of FDA labeling indications (i.e., “off label”) typically requires a physician’s prescription, which will almost certainly reduce COVID-19 vaccine uptake in children and pregnant patients, given that most people received these shots in pharmacies during the 2024-25 season. CVS and Walgreens, the country’s two largest pharmacy chains, are requiring physician prescriptions or waiting for ACIP guidance to make the new vaccines available in many states. However, an increasing number of states have implemented executive orders or passed legislation to permit pharmacists to provide vaccines to anyone who wants them. For example, the Pennsylvania State Board of Pharmacy voted unanimously to issue guidance that would allow pharmacists to administer any vaccines recommended by AAFP, AAP, or ACOG.

Erosion of vaccine uptake could easily worsen the burden of illness for our patients and the health system. Navigating the unnecessarily complex landscape of COVID-19 vaccines will be challenging, but it remains worthwhile.
 

Kenneth W. Lin, MD, MPH, Associate Director, Department of Family Medicine, Lancaster General Hospital, Lancaster, Pennsylvania, has disclosed the following relevant financial relationships: Serve(d) as a director, officer, partner, employee, advisor, consultant, or trustee for: UpToDate; American Academy of Family Physicians; Archdiocese of Washington; Association of Prevention Teaching and Research.

A version of this article appeared on Medscape.com.

Hi, everyone. I’m Dr Kenny Lin. I am a family physician and associate director of the Lancaster General Hospital Family Medicine Residency, and I blog at Common Sense Family Doctor.

The receding of the pandemic and the understandable desire to return to normalcy has made COVID-19 vaccines a lower priority for many of our patients. However, family physicians should keep in mind that from October 1, 2024, to September 6, 2025, COVID-19 was responsible for an estimated 3.2 to 4.6 million outpatient visits, 360,000 to 520,000 hospitalizations, and 42,000 to 60,000 deaths.

In a previous commentary, I discussed the worsening disconnect between the evidence supporting the effectiveness and safety of vaccinations and increasing reluctance of patients and parents to receive them, fueled by misinformation from federal health agencies and the packing of the Advisory Committee on Immunization Practices (ACIP) with vaccine skeptics. Since then, Secretary of Health and Human Services (HHS), Robert F. Kennedy, Jr, has fired Dr Susan Monarez, his handpicked director of the CDC. This caused three senior CDC officials to resign in protest and precipitated further turmoil at the embattled agency. 

The FDA has approved 3 updated COVID-19 vaccines targeted to currently circulating strains: an mRNA vaccine from Moderna (Spikevax) for those aged 6 months or older; an mRNA vaccine from Pfizer/BioNTech (Comirnaty) for those aged ≥ 5 years; and a protein subunit vaccine from Novavax (Nuvaxovid) for those aged ≥ 12 years. However, approvals restricting the scope of these approvals to certain high-risk groups, combined with the ACIP’s recent decision to not explicitly recommend them for any group, have complicated access for many patients.

Medical groups, including the American Academy of Pediatrics (AAP), the American Academy of Family Physicians (AAFP), and the American College of Obstetricians and Gynecologists (ACOG), have published their own recommendations (Table). Of note, in opposition to the FDA and ACIP, the AAP and AAFP strongly recommend routine vaccination for children aged 6 to 23 months because they have the highest risk for hospitalization. The AAFP and ACOG both recommend COVID-19 vaccination in pregnancy to protect the pregnant patient and provide passive antibody protection to their infants up to 6 months of age. The Vaccine Integrity Project’s review of 12 safety studies published since June 2024 found that mRNA vaccines were not associated with increases in any adverse maternal or infant outcomes and had a possible protective effect against preterm birth.

In my previous commentary, 70% of Medscape readers indicated that they would follow vaccination recommendations from AAP even if they differed from CDC guidance. Administering vaccines outside of FDA labeling indications (i.e., “off label”) typically requires a physician’s prescription, which will almost certainly reduce COVID-19 vaccine uptake in children and pregnant patients, given that most people received these shots in pharmacies during the 2024-25 season. CVS and Walgreens, the country’s two largest pharmacy chains, are requiring physician prescriptions or waiting for ACIP guidance to make the new vaccines available in many states. However, an increasing number of states have implemented executive orders or passed legislation to permit pharmacists to provide vaccines to anyone who wants them. For example, the Pennsylvania State Board of Pharmacy voted unanimously to issue guidance that would allow pharmacists to administer any vaccines recommended by AAFP, AAP, or ACOG.

Erosion of vaccine uptake could easily worsen the burden of illness for our patients and the health system. Navigating the unnecessarily complex landscape of COVID-19 vaccines will be challenging, but it remains worthwhile.
 

Kenneth W. Lin, MD, MPH, Associate Director, Department of Family Medicine, Lancaster General Hospital, Lancaster, Pennsylvania, has disclosed the following relevant financial relationships: Serve(d) as a director, officer, partner, employee, advisor, consultant, or trustee for: UpToDate; American Academy of Family Physicians; Archdiocese of Washington; Association of Prevention Teaching and Research.

A version of this article appeared on Medscape.com.

Hi, everyone. I’m Dr Kenny Lin. I am a family physician and associate director of the Lancaster General Hospital Family Medicine Residency, and I blog at Common Sense Family Doctor.

The receding of the pandemic and the understandable desire to return to normalcy has made COVID-19 vaccines a lower priority for many of our patients. However, family physicians should keep in mind that from October 1, 2024, to September 6, 2025, COVID-19 was responsible for an estimated 3.2 to 4.6 million outpatient visits, 360,000 to 520,000 hospitalizations, and 42,000 to 60,000 deaths.

In a previous commentary, I discussed the worsening disconnect between the evidence supporting the effectiveness and safety of vaccinations and increasing reluctance of patients and parents to receive them, fueled by misinformation from federal health agencies and the packing of the Advisory Committee on Immunization Practices (ACIP) with vaccine skeptics. Since then, Secretary of Health and Human Services (HHS), Robert F. Kennedy, Jr, has fired Dr Susan Monarez, his handpicked director of the CDC. This caused three senior CDC officials to resign in protest and precipitated further turmoil at the embattled agency. 

The FDA has approved 3 updated COVID-19 vaccines targeted to currently circulating strains: an mRNA vaccine from Moderna (Spikevax) for those aged 6 months or older; an mRNA vaccine from Pfizer/BioNTech (Comirnaty) for those aged ≥ 5 years; and a protein subunit vaccine from Novavax (Nuvaxovid) for those aged ≥ 12 years. However, approvals restricting the scope of these approvals to certain high-risk groups, combined with the ACIP’s recent decision to not explicitly recommend them for any group, have complicated access for many patients.

Medical groups, including the American Academy of Pediatrics (AAP), the American Academy of Family Physicians (AAFP), and the American College of Obstetricians and Gynecologists (ACOG), have published their own recommendations (Table). Of note, in opposition to the FDA and ACIP, the AAP and AAFP strongly recommend routine vaccination for children aged 6 to 23 months because they have the highest risk for hospitalization. The AAFP and ACOG both recommend COVID-19 vaccination in pregnancy to protect the pregnant patient and provide passive antibody protection to their infants up to 6 months of age. The Vaccine Integrity Project’s review of 12 safety studies published since June 2024 found that mRNA vaccines were not associated with increases in any adverse maternal or infant outcomes and had a possible protective effect against preterm birth.

In my previous commentary, 70% of Medscape readers indicated that they would follow vaccination recommendations from AAP even if they differed from CDC guidance. Administering vaccines outside of FDA labeling indications (i.e., “off label”) typically requires a physician’s prescription, which will almost certainly reduce COVID-19 vaccine uptake in children and pregnant patients, given that most people received these shots in pharmacies during the 2024-25 season. CVS and Walgreens, the country’s two largest pharmacy chains, are requiring physician prescriptions or waiting for ACIP guidance to make the new vaccines available in many states. However, an increasing number of states have implemented executive orders or passed legislation to permit pharmacists to provide vaccines to anyone who wants them. For example, the Pennsylvania State Board of Pharmacy voted unanimously to issue guidance that would allow pharmacists to administer any vaccines recommended by AAFP, AAP, or ACOG.

Erosion of vaccine uptake could easily worsen the burden of illness for our patients and the health system. Navigating the unnecessarily complex landscape of COVID-19 vaccines will be challenging, but it remains worthwhile.
 

Kenneth W. Lin, MD, MPH, Associate Director, Department of Family Medicine, Lancaster General Hospital, Lancaster, Pennsylvania, has disclosed the following relevant financial relationships: Serve(d) as a director, officer, partner, employee, advisor, consultant, or trustee for: UpToDate; American Academy of Family Physicians; Archdiocese of Washington; Association of Prevention Teaching and Research.

A version of this article appeared on Medscape.com.

In this Practical AI column, we’ve explored everything from large language models to the nuances of trial matching, but one of the most immediate and impactful applications of AI is unfolding right now in breast imaging. For oncologists, this isn’t an abstract future — with new screening guidelines, dense-breast mandates, and a shrinking radiology workforce, it’s the imaging reports and patient questions landing in your clinic today.

Here is what oncologists need to know, and how to put it to work for their patients.

 

Why AI in Mammography Matters

More than 200 million women undergo breast cancer screening each year. In the US alone, 10% of the 40 million women screened annually require additional diagnostic imaging, and 4%–5% of these women are eventually diagnosed with breast cancer.

Two major shifts are redefining breast cancer screening in the US: The US Preventive Services Task Force (USPSTF) now recommends biennial screening from age 40 to 74 years, and notifying patients of breast density is a federal requirement as of September 10, 2024. That means more mammograms, more patient questions, and more downstream oncology decisions. Patients will increasingly ask about “dense” breast results and what to do next. Add a national radiologist shortage into the mix, and the pressure on timely callbacks, biopsies, and treatment planning will only grow.

 

Can AI Help Without Compromising Care?

The short answer is yes. With AI, we may be able to transform these rate-limiting steps into opportunities for earlier detection, decentralized screening, and smarter triage and save hundreds of thousands of women from an unnecessary diagnostic procedure, if implemented deliberately.

Don’t Confuse Today’s AI With Yesterday’s CAD 

Think of older computer-aided detection (CAD) like a 1990s chemotherapy drug: It sometimes helped, but it came with significant toxicity and rarely delivered consistent survival benefits. Today’s deep-learning AI is closer to targeted therapy — trained on millions of “trial participants” (mammograms), more precise, and applied in specific contexts where it adds value. If you once dismissed CAD as noise, it’s time to revisit what AI can now offer.

The role of AI is broader than drawing boxes. It provides second readings, worklist triage, risk prediction, density assessment, and decision support. FDA has cleared several AI tools for both 2D and digital breast tomosynthesis (DBT), which include iCAD ProFound (DBT), ScreenPoint Transpara (2D/DBT), and Lunit INSIGHT DBT. 

Some of the strongest evidence for AI in mammography is as a second reader during screening. Large trials show that AI plus one radiologist can match reading from two radiologists, cutting workload by about 40%. For example, the MASAI randomized trial showed that AI-supported screening achieved similar cancer detection but cut human screen-reading workload about 44% vs standard double reading (39,996 vs 40,024 participants). The primary interval cancer outcomes are maturing, but the safety analysis is reassuring.

Reducing second reads and arbitration time are important for clinicians because it frees capacity for callbacks and diagnostic workups. This will be especially key given that screening now starts at age 40. That will mean about 21 to 22 million more women are newly eligible, translating to about 10 to 11 million additional mammograms each year under biennial screening.

Another important area where AI can make its mark in mammography is triage and time to diagnosis. The results from a randomized implementation study showed that AI-prioritized worklists accelerated time to additional imaging and biopsy diagnosis without harming efficiency for others — exactly the kind of outcome patients feel.

Multiple studies have demonstrated improved diagnostic performance and shorter reading times when AI supports DBT interpretation, which is important because DBT can otherwise be time intensive.

We are also seeing rapid advancement in risk-based screening, moving beyond a single dense vs not dense approach. Deep-learning risk models, such as Mirai, predict 1- to 5-year breast cancer risk directly from the mammogram, and these tools are now being assessed prospectively to guide supplemental MRI. Cost-effectiveness modeling supports risk-stratified intervals vs one-size-fits-all schedules.

Finally, automated density tools, such as Transpara Density and Volpara, offer objective, reproducible volumetric measures that map to the Breast Imaging-Reporting and Data System, which is useful for Mammography Quality Standards Act-required reporting and as inputs to risk calculators.

While early evidence suggests AI may help surface future or interval cancers earlier, including more invasive tumors, the definitive impacts on interval cancer rates and mortality require longitudinal follow-up, which is now in progress.

 

Pitfalls to Watch For

Bias is real. Studies show false-positive differences by race, age, and density. AI can even infer racial identity from images, potentially amplifying disparities. Performance can also shift by vendor, demographics, and prevalence.

A Radiology study of 4855 DBT exams showed that an algorithm produced more false-positive case scores in Black patients and older patients (aged 71-80 years) patients and in women with extremely dense breasts. This can happen because AI can infer proxies for race directly from images, even when humans cannot, and this can propagate disparities if not addressed. External validations and reviews emphasize that performance can shift with device manufacturer, demographics, and prevalence, which is why all tools need to undergo local validation and calibration. 

Here’s a pragmatic adoption checklist before going live with an AI tool.

• Confirm FDA clearance: Verify the name and version of the algorithm, imaging modes (2D vs DBT), and operating points. Confirm 510(k) numbers.
• Local validation: Test on your patient mix and vendor stack (Hologic, GE, Siemens, Fuji). Compare this to your baseline recall rate, positive predictive value of recall (PPV1), cancer detection rate, and reading time. Commit to recalibration if drift occurs.
• Equity plan: Monitor false-positive and negative false-rates by age, race/ethnicity, and density; document corrective actions if disparities emerge. (This isn’t optional.)
• Workflow clarity: Is AI a second reader, an additional reader, or a triage tool? Who arbitrates discordance? What’s the escalation path for high-risk or interval cancer-like patterns?
• Regulatory strategy: Confirm whether the vendor has (or will file) a Predetermined Change Control Plan so models can be updated safely without repeated submissions. Also confirm how you’ll be notified about performance-relevant changes.
• Data governance: Audit logs of AI outputs, retention, protected health information handling, and the patient communication policy for AI-assisted reads.

After going live, set up a quarterly dashboard. It should include cancer detection rate per 1000 patients, recall rate, PPV1, interval cancer rate (as it matures), reading time, and turnaround time to diagnostic imaging or biopsy — all stratified by age, race/ethnicity, and density.

Here, I dissect what this discussion means through the lens of Moravec’s paradox (machines excel at what clinicians find hard, and vice versa) and offer a possible playbook for putting these tools to work.

 

What to Tell Patients

When speaking with patients, emphasize that a radiologist still reads their mammogram. AI helps with consistency and efficiency; it doesn’t replace human oversight. Patients with dense breasts should still expect a standard notice; discussion of individualized risk factors, such as family history, genetics, and prior biopsies; and consideration of supplemental imaging if risk warrants. But it’s also important to tell these patients that while dense breasts are common, they do not automatically mean high cancer risk.

As for screening schedules, remind patients that screening is at least biennial from 40 to 74 years of age per the USPSTF guidelines; however, specialty groups may recommend starting on an annual schedule at 40.

 

What You Can Implement Now

There are multiple practical use cases you can introduce now. One is to use AI as a second reader or an additional reader safety net to preserve detection while reducing human workload. This helps your breast center absorb screening expansion to age 40 without diluting quality. Another is to turn on AI triage to shorten the time to callback and biopsy for the few who need it most — patients notice and appreciate faster answers. You can also begin adopting automated density plus risk models to move beyond “dense/not dense.” For selected patients, AI-informed risk can justify MRI or tailored intervals. 

Here’s a quick cheat sheet (for your next leadership or tumor-board meeting).

 

Do:

• Use AI as a second or additional reader or triage tool, not as a black box.
• Track cancer detection rate, recall, PPV1, interval cancers, and reading time, stratified by age, race, and breast density.
• Pair automated density with AI risk to personalize screening and supplemental imaging.
• Enroll patients in future clinical trials, such as PRISM, the first large-scale randomized controlled trial of AI for screening mammography. This US-based, $16 million, seven-site study is funded by the Patient-Centered Outcomes Research Institute.

Don’t:

• Assume “AI = CAD.” The 2015 CAD story is over; modern deep learning systems are different and require different oversight.
• Go live without a local validation and equity plan or without clarity on software updates.
• Forget to remind patients that screening starts at age 40, and dense breast notifications are now universal. Use the visit to discuss risk, supplemental imaging, and why a human still directs their care.

The Bottom Line

AI won’t replace radiologists or read mammograms for us — just as PET scans didn’t replace oncologists and stethoscopes didn’t make cardiologists obsolete. What it will do is catch what the tired human eye might miss, shave days off anxious waiting, and turn breast density into data instead of doubt. For oncologists, that means staging sooner, enrolling smarter, and spending more time talking with patients instead of chasing callbacks.

In short, AI may not take the picture, but it helps us frame the story, making it sharper, faster, and with fewer blind spots. By pairing this powerful technology with rigorous, equity-focused local validation and transparent governance under the FDA’s emerging Predetermined Change Control Plan framework, we can realize the tangible benefits of practical AI for our patients without widening disparities. 

Now, during Breast Cancer Awareness Month, how about we add on AI to that pink ribbon — how cool would that be?

Thoughts? Drop me a line at [email protected]. Let’s keep the conversation — and pink ribbons — going.

Arturo Loaiza-Bonilla, MD, MSEd, is the co-founder and chief medical AI officer at Massive Bio, a company connecting patients to clinical trials using artificial intelligence. His research and professional interests focus on precision medicine, clinical trial design, digital health, entrepreneurship, and patient advocacy. Dr Loaiza-Bonilla serves as Systemwide Chief of Hematology and Oncology at St. Luke’s University Health Network, where he maintains a connection to patient care by attending to patients 2 days a week.

A version of this article first appeared on Medscape.com.

In this Practical AI column, we’ve explored everything from large language models to the nuances of trial matching, but one of the most immediate and impactful applications of AI is unfolding right now in breast imaging. For oncologists, this isn’t an abstract future — with new screening guidelines, dense-breast mandates, and a shrinking radiology workforce, it’s the imaging reports and patient questions landing in your clinic today.

Here is what oncologists need to know, and how to put it to work for their patients.

 

Why AI in Mammography Matters

More than 200 million women undergo breast cancer screening each year. In the US alone, 10% of the 40 million women screened annually require additional diagnostic imaging, and 4%–5% of these women are eventually diagnosed with breast cancer.

Two major shifts are redefining breast cancer screening in the US: The US Preventive Services Task Force (USPSTF) now recommends biennial screening from age 40 to 74 years, and notifying patients of breast density is a federal requirement as of September 10, 2024. That means more mammograms, more patient questions, and more downstream oncology decisions. Patients will increasingly ask about “dense” breast results and what to do next. Add a national radiologist shortage into the mix, and the pressure on timely callbacks, biopsies, and treatment planning will only grow.

 

Can AI Help Without Compromising Care?

The short answer is yes. With AI, we may be able to transform these rate-limiting steps into opportunities for earlier detection, decentralized screening, and smarter triage and save hundreds of thousands of women from an unnecessary diagnostic procedure, if implemented deliberately.

Don’t Confuse Today’s AI With Yesterday’s CAD 

Think of older computer-aided detection (CAD) like a 1990s chemotherapy drug: It sometimes helped, but it came with significant toxicity and rarely delivered consistent survival benefits. Today’s deep-learning AI is closer to targeted therapy — trained on millions of “trial participants” (mammograms), more precise, and applied in specific contexts where it adds value. If you once dismissed CAD as noise, it’s time to revisit what AI can now offer.

The role of AI is broader than drawing boxes. It provides second readings, worklist triage, risk prediction, density assessment, and decision support. FDA has cleared several AI tools for both 2D and digital breast tomosynthesis (DBT), which include iCAD ProFound (DBT), ScreenPoint Transpara (2D/DBT), and Lunit INSIGHT DBT. 

Some of the strongest evidence for AI in mammography is as a second reader during screening. Large trials show that AI plus one radiologist can match reading from two radiologists, cutting workload by about 40%. For example, the MASAI randomized trial showed that AI-supported screening achieved similar cancer detection but cut human screen-reading workload about 44% vs standard double reading (39,996 vs 40,024 participants). The primary interval cancer outcomes are maturing, but the safety analysis is reassuring.

Reducing second reads and arbitration time are important for clinicians because it frees capacity for callbacks and diagnostic workups. This will be especially key given that screening now starts at age 40. That will mean about 21 to 22 million more women are newly eligible, translating to about 10 to 11 million additional mammograms each year under biennial screening.

Another important area where AI can make its mark in mammography is triage and time to diagnosis. The results from a randomized implementation study showed that AI-prioritized worklists accelerated time to additional imaging and biopsy diagnosis without harming efficiency for others — exactly the kind of outcome patients feel.

Multiple studies have demonstrated improved diagnostic performance and shorter reading times when AI supports DBT interpretation, which is important because DBT can otherwise be time intensive.

We are also seeing rapid advancement in risk-based screening, moving beyond a single dense vs not dense approach. Deep-learning risk models, such as Mirai, predict 1- to 5-year breast cancer risk directly from the mammogram, and these tools are now being assessed prospectively to guide supplemental MRI. Cost-effectiveness modeling supports risk-stratified intervals vs one-size-fits-all schedules.

Finally, automated density tools, such as Transpara Density and Volpara, offer objective, reproducible volumetric measures that map to the Breast Imaging-Reporting and Data System, which is useful for Mammography Quality Standards Act-required reporting and as inputs to risk calculators.

While early evidence suggests AI may help surface future or interval cancers earlier, including more invasive tumors, the definitive impacts on interval cancer rates and mortality require longitudinal follow-up, which is now in progress.

 

Pitfalls to Watch For

Bias is real. Studies show false-positive differences by race, age, and density. AI can even infer racial identity from images, potentially amplifying disparities. Performance can also shift by vendor, demographics, and prevalence.

A Radiology study of 4855 DBT exams showed that an algorithm produced more false-positive case scores in Black patients and older patients (aged 71-80 years) patients and in women with extremely dense breasts. This can happen because AI can infer proxies for race directly from images, even when humans cannot, and this can propagate disparities if not addressed. External validations and reviews emphasize that performance can shift with device manufacturer, demographics, and prevalence, which is why all tools need to undergo local validation and calibration. 

Here’s a pragmatic adoption checklist before going live with an AI tool.

• Confirm FDA clearance: Verify the name and version of the algorithm, imaging modes (2D vs DBT), and operating points. Confirm 510(k) numbers.
• Local validation: Test on your patient mix and vendor stack (Hologic, GE, Siemens, Fuji). Compare this to your baseline recall rate, positive predictive value of recall (PPV1), cancer detection rate, and reading time. Commit to recalibration if drift occurs.
• Equity plan: Monitor false-positive and negative false-rates by age, race/ethnicity, and density; document corrective actions if disparities emerge. (This isn’t optional.)
• Workflow clarity: Is AI a second reader, an additional reader, or a triage tool? Who arbitrates discordance? What’s the escalation path for high-risk or interval cancer-like patterns?
• Regulatory strategy: Confirm whether the vendor has (or will file) a Predetermined Change Control Plan so models can be updated safely without repeated submissions. Also confirm how you’ll be notified about performance-relevant changes.
• Data governance: Audit logs of AI outputs, retention, protected health information handling, and the patient communication policy for AI-assisted reads.

After going live, set up a quarterly dashboard. It should include cancer detection rate per 1000 patients, recall rate, PPV1, interval cancer rate (as it matures), reading time, and turnaround time to diagnostic imaging or biopsy — all stratified by age, race/ethnicity, and density.

Here, I dissect what this discussion means through the lens of Moravec’s paradox (machines excel at what clinicians find hard, and vice versa) and offer a possible playbook for putting these tools to work.

 

What to Tell Patients

When speaking with patients, emphasize that a radiologist still reads their mammogram. AI helps with consistency and efficiency; it doesn’t replace human oversight. Patients with dense breasts should still expect a standard notice; discussion of individualized risk factors, such as family history, genetics, and prior biopsies; and consideration of supplemental imaging if risk warrants. But it’s also important to tell these patients that while dense breasts are common, they do not automatically mean high cancer risk.

As for screening schedules, remind patients that screening is at least biennial from 40 to 74 years of age per the USPSTF guidelines; however, specialty groups may recommend starting on an annual schedule at 40.

 

What You Can Implement Now

There are multiple practical use cases you can introduce now. One is to use AI as a second reader or an additional reader safety net to preserve detection while reducing human workload. This helps your breast center absorb screening expansion to age 40 without diluting quality. Another is to turn on AI triage to shorten the time to callback and biopsy for the few who need it most — patients notice and appreciate faster answers. You can also begin adopting automated density plus risk models to move beyond “dense/not dense.” For selected patients, AI-informed risk can justify MRI or tailored intervals. 

Here’s a quick cheat sheet (for your next leadership or tumor-board meeting).

 

Do:

• Use AI as a second or additional reader or triage tool, not as a black box.
• Track cancer detection rate, recall, PPV1, interval cancers, and reading time, stratified by age, race, and breast density.
• Pair automated density with AI risk to personalize screening and supplemental imaging.
• Enroll patients in future clinical trials, such as PRISM, the first large-scale randomized controlled trial of AI for screening mammography. This US-based, $16 million, seven-site study is funded by the Patient-Centered Outcomes Research Institute.

Don’t:

• Assume “AI = CAD.” The 2015 CAD story is over; modern deep learning systems are different and require different oversight.
• Go live without a local validation and equity plan or without clarity on software updates.
• Forget to remind patients that screening starts at age 40, and dense breast notifications are now universal. Use the visit to discuss risk, supplemental imaging, and why a human still directs their care.

The Bottom Line

AI won’t replace radiologists or read mammograms for us — just as PET scans didn’t replace oncologists and stethoscopes didn’t make cardiologists obsolete. What it will do is catch what the tired human eye might miss, shave days off anxious waiting, and turn breast density into data instead of doubt. For oncologists, that means staging sooner, enrolling smarter, and spending more time talking with patients instead of chasing callbacks.

In short, AI may not take the picture, but it helps us frame the story, making it sharper, faster, and with fewer blind spots. By pairing this powerful technology with rigorous, equity-focused local validation and transparent governance under the FDA’s emerging Predetermined Change Control Plan framework, we can realize the tangible benefits of practical AI for our patients without widening disparities. 

Now, during Breast Cancer Awareness Month, how about we add on AI to that pink ribbon — how cool would that be?

Thoughts? Drop me a line at [email protected]. Let’s keep the conversation — and pink ribbons — going.

Arturo Loaiza-Bonilla, MD, MSEd, is the co-founder and chief medical AI officer at Massive Bio, a company connecting patients to clinical trials using artificial intelligence. His research and professional interests focus on precision medicine, clinical trial design, digital health, entrepreneurship, and patient advocacy. Dr Loaiza-Bonilla serves as Systemwide Chief of Hematology and Oncology at St. Luke’s University Health Network, where he maintains a connection to patient care by attending to patients 2 days a week.

A version of this article first appeared on Medscape.com.

In this Practical AI column, we’ve explored everything from large language models to the nuances of trial matching, but one of the most immediate and impactful applications of AI is unfolding right now in breast imaging. For oncologists, this isn’t an abstract future — with new screening guidelines, dense-breast mandates, and a shrinking radiology workforce, it’s the imaging reports and patient questions landing in your clinic today.

Here is what oncologists need to know, and how to put it to work for their patients.

 

Why AI in Mammography Matters

More than 200 million women undergo breast cancer screening each year. In the US alone, 10% of the 40 million women screened annually require additional diagnostic imaging, and 4%–5% of these women are eventually diagnosed with breast cancer.

Two major shifts are redefining breast cancer screening in the US: The US Preventive Services Task Force (USPSTF) now recommends biennial screening from age 40 to 74 years, and notifying patients of breast density is a federal requirement as of September 10, 2024. That means more mammograms, more patient questions, and more downstream oncology decisions. Patients will increasingly ask about “dense” breast results and what to do next. Add a national radiologist shortage into the mix, and the pressure on timely callbacks, biopsies, and treatment planning will only grow.

 

Can AI Help Without Compromising Care?

The short answer is yes. With AI, we may be able to transform these rate-limiting steps into opportunities for earlier detection, decentralized screening, and smarter triage and save hundreds of thousands of women from an unnecessary diagnostic procedure, if implemented deliberately.

Don’t Confuse Today’s AI With Yesterday’s CAD 

Think of older computer-aided detection (CAD) like a 1990s chemotherapy drug: It sometimes helped, but it came with significant toxicity and rarely delivered consistent survival benefits. Today’s deep-learning AI is closer to targeted therapy — trained on millions of “trial participants” (mammograms), more precise, and applied in specific contexts where it adds value. If you once dismissed CAD as noise, it’s time to revisit what AI can now offer.

The role of AI is broader than drawing boxes. It provides second readings, worklist triage, risk prediction, density assessment, and decision support. FDA has cleared several AI tools for both 2D and digital breast tomosynthesis (DBT), which include iCAD ProFound (DBT), ScreenPoint Transpara (2D/DBT), and Lunit INSIGHT DBT. 

Some of the strongest evidence for AI in mammography is as a second reader during screening. Large trials show that AI plus one radiologist can match reading from two radiologists, cutting workload by about 40%. For example, the MASAI randomized trial showed that AI-supported screening achieved similar cancer detection but cut human screen-reading workload about 44% vs standard double reading (39,996 vs 40,024 participants). The primary interval cancer outcomes are maturing, but the safety analysis is reassuring.

Reducing second reads and arbitration time are important for clinicians because it frees capacity for callbacks and diagnostic workups. This will be especially key given that screening now starts at age 40. That will mean about 21 to 22 million more women are newly eligible, translating to about 10 to 11 million additional mammograms each year under biennial screening.

Another important area where AI can make its mark in mammography is triage and time to diagnosis. The results from a randomized implementation study showed that AI-prioritized worklists accelerated time to additional imaging and biopsy diagnosis without harming efficiency for others — exactly the kind of outcome patients feel.

Multiple studies have demonstrated improved diagnostic performance and shorter reading times when AI supports DBT interpretation, which is important because DBT can otherwise be time intensive.

We are also seeing rapid advancement in risk-based screening, moving beyond a single dense vs not dense approach. Deep-learning risk models, such as Mirai, predict 1- to 5-year breast cancer risk directly from the mammogram, and these tools are now being assessed prospectively to guide supplemental MRI. Cost-effectiveness modeling supports risk-stratified intervals vs one-size-fits-all schedules.

Finally, automated density tools, such as Transpara Density and Volpara, offer objective, reproducible volumetric measures that map to the Breast Imaging-Reporting and Data System, which is useful for Mammography Quality Standards Act-required reporting and as inputs to risk calculators.

While early evidence suggests AI may help surface future or interval cancers earlier, including more invasive tumors, the definitive impacts on interval cancer rates and mortality require longitudinal follow-up, which is now in progress.

 

Pitfalls to Watch For

Bias is real. Studies show false-positive differences by race, age, and density. AI can even infer racial identity from images, potentially amplifying disparities. Performance can also shift by vendor, demographics, and prevalence.

A Radiology study of 4855 DBT exams showed that an algorithm produced more false-positive case scores in Black patients and older patients (aged 71-80 years) patients and in women with extremely dense breasts. This can happen because AI can infer proxies for race directly from images, even when humans cannot, and this can propagate disparities if not addressed. External validations and reviews emphasize that performance can shift with device manufacturer, demographics, and prevalence, which is why all tools need to undergo local validation and calibration. 

Here’s a pragmatic adoption checklist before going live with an AI tool.

• Confirm FDA clearance: Verify the name and version of the algorithm, imaging modes (2D vs DBT), and operating points. Confirm 510(k) numbers.
• Local validation: Test on your patient mix and vendor stack (Hologic, GE, Siemens, Fuji). Compare this to your baseline recall rate, positive predictive value of recall (PPV1), cancer detection rate, and reading time. Commit to recalibration if drift occurs.
• Equity plan: Monitor false-positive and negative false-rates by age, race/ethnicity, and density; document corrective actions if disparities emerge. (This isn’t optional.)
• Workflow clarity: Is AI a second reader, an additional reader, or a triage tool? Who arbitrates discordance? What’s the escalation path for high-risk or interval cancer-like patterns?
• Regulatory strategy: Confirm whether the vendor has (or will file) a Predetermined Change Control Plan so models can be updated safely without repeated submissions. Also confirm how you’ll be notified about performance-relevant changes.
• Data governance: Audit logs of AI outputs, retention, protected health information handling, and the patient communication policy for AI-assisted reads.

After going live, set up a quarterly dashboard. It should include cancer detection rate per 1000 patients, recall rate, PPV1, interval cancer rate (as it matures), reading time, and turnaround time to diagnostic imaging or biopsy — all stratified by age, race/ethnicity, and density.

Here, I dissect what this discussion means through the lens of Moravec’s paradox (machines excel at what clinicians find hard, and vice versa) and offer a possible playbook for putting these tools to work.

 

What to Tell Patients

When speaking with patients, emphasize that a radiologist still reads their mammogram. AI helps with consistency and efficiency; it doesn’t replace human oversight. Patients with dense breasts should still expect a standard notice; discussion of individualized risk factors, such as family history, genetics, and prior biopsies; and consideration of supplemental imaging if risk warrants. But it’s also important to tell these patients that while dense breasts are common, they do not automatically mean high cancer risk.

As for screening schedules, remind patients that screening is at least biennial from 40 to 74 years of age per the USPSTF guidelines; however, specialty groups may recommend starting on an annual schedule at 40.

 

What You Can Implement Now

There are multiple practical use cases you can introduce now. One is to use AI as a second reader or an additional reader safety net to preserve detection while reducing human workload. This helps your breast center absorb screening expansion to age 40 without diluting quality. Another is to turn on AI triage to shorten the time to callback and biopsy for the few who need it most — patients notice and appreciate faster answers. You can also begin adopting automated density plus risk models to move beyond “dense/not dense.” For selected patients, AI-informed risk can justify MRI or tailored intervals. 

Here’s a quick cheat sheet (for your next leadership or tumor-board meeting).

 

Do:

• Use AI as a second or additional reader or triage tool, not as a black box.
• Track cancer detection rate, recall, PPV1, interval cancers, and reading time, stratified by age, race, and breast density.
• Pair automated density with AI risk to personalize screening and supplemental imaging.
• Enroll patients in future clinical trials, such as PRISM, the first large-scale randomized controlled trial of AI for screening mammography. This US-based, $16 million, seven-site study is funded by the Patient-Centered Outcomes Research Institute.

Don’t:

• Assume “AI = CAD.” The 2015 CAD story is over; modern deep learning systems are different and require different oversight.
• Go live without a local validation and equity plan or without clarity on software updates.
• Forget to remind patients that screening starts at age 40, and dense breast notifications are now universal. Use the visit to discuss risk, supplemental imaging, and why a human still directs their care.

The Bottom Line

AI won’t replace radiologists or read mammograms for us — just as PET scans didn’t replace oncologists and stethoscopes didn’t make cardiologists obsolete. What it will do is catch what the tired human eye might miss, shave days off anxious waiting, and turn breast density into data instead of doubt. For oncologists, that means staging sooner, enrolling smarter, and spending more time talking with patients instead of chasing callbacks.

In short, AI may not take the picture, but it helps us frame the story, making it sharper, faster, and with fewer blind spots. By pairing this powerful technology with rigorous, equity-focused local validation and transparent governance under the FDA’s emerging Predetermined Change Control Plan framework, we can realize the tangible benefits of practical AI for our patients without widening disparities. 

Now, during Breast Cancer Awareness Month, how about we add on AI to that pink ribbon — how cool would that be?

Thoughts? Drop me a line at [email protected]. Let’s keep the conversation — and pink ribbons — going.

Arturo Loaiza-Bonilla, MD, MSEd, is the co-founder and chief medical AI officer at Massive Bio, a company connecting patients to clinical trials using artificial intelligence. His research and professional interests focus on precision medicine, clinical trial design, digital health, entrepreneurship, and patient advocacy. Dr Loaiza-Bonilla serves as Systemwide Chief of Hematology and Oncology at St. Luke’s University Health Network, where he maintains a connection to patient care by attending to patients 2 days a week.

A version of this article first appeared on Medscape.com.

Last month, I had the privilege of joining more than one hundred physician colleagues in Washington, DC, for AGA Advocacy Day. While standing amidst the majesty of the Capital, I found myself deeply appreciative for those who dedicate their time and energy to public service. Many of these dedicated federal workers choose to be in DC because of a sincere belief in their mission.

Among these mission-driven public servants are federal employees who work in the Department of Veterans Affairs (VA). As a member of this group, I come to work energized by the mission to care for those who have served in our military. In my clinical practice, I am reminded regularly of the sacrifices of veterans and their families. This month, and especially on Veterans Day, I hope we will take a moment to express gratitude to veterans for their service to our country.

This month is also a timely opportunity to recognize the immense contributions of VA physicians to the field of gastroenterology. Many young gastroenterologists may not know that it was the landmark VA Cooperative Study #380, led by Dr. David Lieberman (Portland VA) that helped push Medicare to cover reimbursement for screening colonoscopy. Today, one of the most important ongoing studies in our field – VA Cooperative Study #577 – continues the VA tradition of high-impact health services research. Launched in 2012, the study has enrolled 50,000 veterans to compare FIT and colonoscopy. It is led by Dr. Jason Dominitz (Seattle VA) and Dr. Doug Robertson (White River Junction VA). 

Beyond research, VA gastroenterologists play a critical role in training the next generation of clinicians. Over 700 gastroenterologists count the VA as a clinical home, making it the largest GI group practice in the country. Many of us — myself included — were trained or mentored by VA physicians whose dedication to service and science has shaped our careers and the field at large.

This month’s issue of GI & Hepatology News has stories about other important contributions to our field. The stories and perspective pieces on Artificial Intelligence are particularly poignant given the announcement last month on the awarding of the Nobel Prize in economics to researchers who study “creative destruction,” the way in which one technological innovation renders others obsolete. Perhaps this award offers another reason to reemphasize and embrace the “art” of medicine.

The views expressed here are my own and do not necessarily reflect the official policy or position of the U.S. Department of Veterans Affairs or the United States Government.

Ziad Gellad, MD, MPH, AGAF

Associate Editor

Last month, I had the privilege of joining more than one hundred physician colleagues in Washington, DC, for AGA Advocacy Day. While standing amidst the majesty of the Capital, I found myself deeply appreciative for those who dedicate their time and energy to public service. Many of these dedicated federal workers choose to be in DC because of a sincere belief in their mission.

Among these mission-driven public servants are federal employees who work in the Department of Veterans Affairs (VA). As a member of this group, I come to work energized by the mission to care for those who have served in our military. In my clinical practice, I am reminded regularly of the sacrifices of veterans and their families. This month, and especially on Veterans Day, I hope we will take a moment to express gratitude to veterans for their service to our country.

This month is also a timely opportunity to recognize the immense contributions of VA physicians to the field of gastroenterology. Many young gastroenterologists may not know that it was the landmark VA Cooperative Study #380, led by Dr. David Lieberman (Portland VA) that helped push Medicare to cover reimbursement for screening colonoscopy. Today, one of the most important ongoing studies in our field – VA Cooperative Study #577 – continues the VA tradition of high-impact health services research. Launched in 2012, the study has enrolled 50,000 veterans to compare FIT and colonoscopy. It is led by Dr. Jason Dominitz (Seattle VA) and Dr. Doug Robertson (White River Junction VA). 

Beyond research, VA gastroenterologists play a critical role in training the next generation of clinicians. Over 700 gastroenterologists count the VA as a clinical home, making it the largest GI group practice in the country. Many of us — myself included — were trained or mentored by VA physicians whose dedication to service and science has shaped our careers and the field at large.

This month’s issue of GI & Hepatology News has stories about other important contributions to our field. The stories and perspective pieces on Artificial Intelligence are particularly poignant given the announcement last month on the awarding of the Nobel Prize in economics to researchers who study “creative destruction,” the way in which one technological innovation renders others obsolete. Perhaps this award offers another reason to reemphasize and embrace the “art” of medicine.

The views expressed here are my own and do not necessarily reflect the official policy or position of the U.S. Department of Veterans Affairs or the United States Government.

Ziad Gellad, MD, MPH, AGAF

Associate Editor

Last month, I had the privilege of joining more than one hundred physician colleagues in Washington, DC, for AGA Advocacy Day. While standing amidst the majesty of the Capital, I found myself deeply appreciative for those who dedicate their time and energy to public service. Many of these dedicated federal workers choose to be in DC because of a sincere belief in their mission.

Among these mission-driven public servants are federal employees who work in the Department of Veterans Affairs (VA). As a member of this group, I come to work energized by the mission to care for those who have served in our military. In my clinical practice, I am reminded regularly of the sacrifices of veterans and their families. This month, and especially on Veterans Day, I hope we will take a moment to express gratitude to veterans for their service to our country.

This month is also a timely opportunity to recognize the immense contributions of VA physicians to the field of gastroenterology. Many young gastroenterologists may not know that it was the landmark VA Cooperative Study #380, led by Dr. David Lieberman (Portland VA) that helped push Medicare to cover reimbursement for screening colonoscopy. Today, one of the most important ongoing studies in our field – VA Cooperative Study #577 – continues the VA tradition of high-impact health services research. Launched in 2012, the study has enrolled 50,000 veterans to compare FIT and colonoscopy. It is led by Dr. Jason Dominitz (Seattle VA) and Dr. Doug Robertson (White River Junction VA). 

Beyond research, VA gastroenterologists play a critical role in training the next generation of clinicians. Over 700 gastroenterologists count the VA as a clinical home, making it the largest GI group practice in the country. Many of us — myself included — were trained or mentored by VA physicians whose dedication to service and science has shaped our careers and the field at large.

This month’s issue of GI & Hepatology News has stories about other important contributions to our field. The stories and perspective pieces on Artificial Intelligence are particularly poignant given the announcement last month on the awarding of the Nobel Prize in economics to researchers who study “creative destruction,” the way in which one technological innovation renders others obsolete. Perhaps this award offers another reason to reemphasize and embrace the “art” of medicine.

The views expressed here are my own and do not necessarily reflect the official policy or position of the U.S. Department of Veterans Affairs or the United States Government.

Ziad Gellad, MD, MPH, AGAF

Associate Editor

Dear colleagues,

Since our last Perspectives feature on artificial intelligence (AI) in gastroenterology and hepatology, the field has experienced remarkable growth in both innovation and clinical adoption. AI tools that were once conceptual are now entering everyday practice, with many more on the horizon poised to transform how we diagnose, treat, and manage patients. In this issue of Perspectives, we present two timely essays that explore how AI is reshaping clinical care—while also emphasizing the need for caution, thoughtful integration, and ongoing oversight.

Dr. Yuvaraj Singh, Dr. Alessandro Colletta, and Dr. Neil Marya discuss how purpose-built AI models can reduce diagnostic uncertainty in advanced endoscopy. From cholangioscopy systems that outperform standard ERCP sampling in distinguishing malignant biliary strictures to EUS-based platforms that differentiate autoimmune pancreatitis from pancreatic cancer, they envision a near-term future in which machine intelligence enhances accuracy, accelerates decision-making, and refines interpretation—without replacing the clinician’s expertise.

Complementing this, Dr. Dennis Shung takes a broader view across the endoscopy unit and outpatient clinic. He highlights the promise of AI for polyp detection, digital biopsy, and automated reporting, while underscoring the importance of human oversight, workflow integration, and safeguards against misinformation. Dr. Shung also emphasizes the pivotal role professional societies can play in establishing clear standards, ethical boundaries, and trusted frameworks for AI deployment in GI practice.

We hope these perspectives spark practical conversations about when—and how—to integrate AI in your own practice. As always, we welcome your feedback and real-world experience. Join the conversation on X at @AGA_GIHN.

Gyanprakash A. Ketwaroo, MD, MSc, is associate professor of medicine, Yale University, New Haven, and chief of endoscopy at West Haven VA Medical Center, both in Connecticut. He is an associate editor for GI & Hepatology News.

AI Models in Advanced Endoscopy

BY YUVARAJ SINGH, MD; ALESSANDRO COLLETTA, MD; NEIL MARYA, MD

As the adage goes, “if tumor is the rumor, then tissue is the issue, because cancer may be the answer.”

Establishing an accurate diagnosis is the essential first step toward curing or palliating malignancy. From detecting an early neoplastic lesion, to distinguishing between malignant and benign pathology, or to determining when and where to obtain tissue, endoscopists are frequently faced with the challenge of transforming diagnostic suspicion into certainty.

Artificial intelligence (AI), designed to replicate human cognition such as pattern recognition and decision-making, has emerged as a technology to assist gastroenterologists in addressing a variety of different tasks during endoscopy. AI research in gastrointestinal endoscopy has initially focused on computer-aided detection (CADe) of colorectal polyps. More recently, however, there has been increased emphasis on developing AI to assist advanced endoscopists.

For instance, in biliary endoscopy, AI is being explored to improve the notoriously challenging diagnosis of cholangiocarcinoma, where conventional tissue sampling often falls short of providing a definitive diagnosis. Similarly, in the pancreas, AI models are showing potential to differentiate autoimmune pancreatitis (AIP) from pancreatic ductal adenocarcinoma (PDAC), a distinction with profound therapeutic implications. Even pancreatic cysts are beginning to benefit from AI models that refine risk stratification and guide management. Together, these advances underscore how AI is not merely an adjunct but a potentially massive catalyst for reimagining the diagnostic role of advanced endoscopists.

Classifying biliary strictures (MBS) accurately remains a challenge. Standard ERCP-based sampling techniques (forceps biopsy and brush cytology) are suboptimal diagnostic tools with false negative rates for detecting MBS of less than 50%. The diagnostic uncertainty related to biliary stricture classification carries significant consequences for patients. For example, patients with biliary cancer without positive cytology have treatments delayed until a malignant diagnosis is established. 

Ancillary technologies to enhance ERCP-based tissue acquisition are still weighed down by low sensitivity and accuracy; even with ancillary use of fluorescent in situ hybridization (FISH), diagnostic yield remains limited. EUS-FNA can help with distal biliary strictures, but this technique risks needle-tract seeding in cases of perihilar disease. Cholangioscopy allows for direct visualization and targeted sampling; however, cholangioscopy-guided forceps biopsies are burdened by low sensitivities.¹ Additionally, physician interpretation of visual findings during cholangioscopy often suffers from poor interobserver agreement and poor accuracy.²

To improve the classification of biliary strictures, several groups have studied the application of AI for cholangioscopy footage of biliary pathology. In our lab, we trained an AI incorporating over 2.3 million cholangioscopy still images and nearly 20,000 expert-annotated frames to enhance its development. The AI closely mirrored expert labeling of cholangioscopy images of malignant pathology and, when tested on full cholangioscopy videos of indeterminate biliary strictures, the AI achieved a diagnostic accuracy of 91%—outperforming both brush cytology (63%) and forceps biopsy (61%).³

The results from this initial study were later validated across multiple centers. AI-assisted cholangioscopy could thus offer a reproducible, real-world solution to one of the most persistent diagnostic dilemmas advanced endoscopists face—helping clinicians act earlier and with greater confidence when evaluating indeterminate strictures.

Moving from the biliary tree to the pancreas, autoimmune pancreatitis (AIP) is a benign fibro-inflammatory disease that often frustrates advanced endoscopists as it closely mimics the appearance of pancreatic ductal adenocarcinoma (PDAC). The stakes are high: despite modern diagnostic techniques, including advanced imaging, some patients with pancreatic resections for “suspected PDAC” are still found to have AIP on final pathology. Conventional tools to distinguish AIP from PDAC have gaps: serum IgG4 and EUS-guided biopsies are both specific but insensitive.

Using EUS videos and images of various pancreas pathologies at Mayo Clinic, we developed an AI to tackle this dilemma. After intensive training, the EUS AI achieved a greater accuracy for distinguishing AIP from PDAC than a group of expert Mayo clinic endosonographers.⁵ In practice, an EUS-AI can identify AIP patterns in real-time, guiding clinicians toward steroid trials or biopsies and reducing the need for unnecessary surgeries.

Looking ahead, there are multiple opportunities for integration of AI into advanced endoscopy practices. Ongoing research suggests that AI could soon assist with identification of pancreas cysts most at risk for malignant transformation, classification of high risk Barrett’s esophagus, and even help with rapid on-site assessment of cytologic specimens obtained during EUS. Beyond diagnosis, AI could likely play an important role in guiding therapeutic interventions. For example, an ERCP AI in the future may be able to provide cannulation assistance or an AI assistant could help endosonographers during deployments of lumen apposing metal stents.

By enhancing image interpretation and procedural consistency, AI has the potential to uphold the fundamental principle of primum non nocere, enabling us to intervene with precision while minimizing harm. AI can also bridge grey zones in clinical practice and narrow diagnostic uncertainty in real time. Importantly, these systems can help clinicians achieve expertise in a fraction of the time it traditionally takes to acquire comparable human proficiency, while offering wider availability across practice settings and reducing interobserver variability that has long challenged endoscopic interpretation.

Currently, adoption is limited by high bias risk, lack of external validation, and interpretability Still, the trajectory of AI suggests a future where these computer technologies will not only support but also elevate human expertise, reshaping the standards of care of diseases managed by advanced endoscopists.

Dr. Singh, Dr. Colletta, and Dr. Marya are based at the Division of Gastroenterology and Hepatology, UMass Chan Medical School, Worcester, Massachusetts. Dr. Marya is a consultant for Boston Scientific, and has no other disclosures. Dr. Singh and Dr. Colletta have no disclosures.

References

1. Navaneethan U, et al. Comparative effectiveness of biliary brush cytology and intraductal biopsy for detection of malignant biliary strictures: a systematic review and meta-analysis. Gastrointest Endosc. 2015 Jan. doi: 10.1016/j.gie.2014.09.017.

2. Stassen PMC, et al. Diagnostic accuracy and interobserver agreement of digital single-operator cholangioscopy for indeterminate biliary strictures. Gastrointest Endosc 2021 Dec. doi: 10.1016/j.gie.2021.06.027.

3. Marya NB, et al. Identification of patients with malignant biliary strictures using a cholangioscopy-based deep learning artificial intelligence (with video). Gastrointest Endosc. 2023 Feb. doi: 10.1016/j.gie.2022.08.021.

4. Marya NB, et al. Multicenter validation of a cholangioscopy artificial intelligence system for the evaluation of biliary tract disease. Endoscopy. 2025 Aug. doi: 10.1055/a-2650-0789.

5. Marya NB, et al. Utilisation of artificial intelligence for the development of an EUS-convolutional neural network model trained to enhance the diagnosis of autoimmune pancreatitis. Gut. 2021 Jul. doi: 10.1136/gutjnl-2020-322821.

AI in General GI and Endoscopy

BY DENNIS L. SHUNG, MD, MHS, PHD

The practice of gastroenterology is changing, but much of it will be rooted in the same – careful, focused attention on endoscopic procedures, and compassionate, attentive care in clinic. Artificial intelligence (AI), like the Industrial Revolution before, is going to transform our practice. This comes with upsides and downsides, and highlights the need for strong leadership from our societies to safeguard the technology for practitioners and patients.

What are the upsides? 

AI has the potential to serve as a second set of eyes in detecting colon polyps, increasing the adenoma detection rate (ADR).¹ AI can be applied to all areas of the gastrointestinal tract, providing digital biopsies, guiding resection, and ensuring quality, which are all now possible with powerful new endoscopy foundation models, such as GastroNet-5M.²

Additionally. the advent of automating the collection of data into reports may herald the end of our days as data entry clerks. Generative AI also has the potential to give us all the best information at our fingertips, suggesting guideline-based care, providing the most up to date evidence, and guiding the differential diagnosis. The potential for patient-facing AI systems could lead to better health literacy, more meaningful engagement, and improved patient satisfaction.³

What are the downsides? 

For endoscopy, AI cannot make up for poor technique to ensure adequate mucosal exposure by the endoscopist, and an increase in AI-supported ADR does not yet convincingly translate into concrete gains in colorectal cancer-related mortality. For the foreseeable future, AI cannot make a connection with the patient in front of us, which is critical in diagnosing and treating patients.

Currently, AI appears to worsen loneliness⁴, and does not necessarily deepen the bonds or provide the positive touch that can heal, and which for many of us, was the reason we became physicians. Finally, as information proliferates, the information risk to patients and providers is growing – in the future, trusted sources to monitor, curate, and guide AI will be ever more important.

 

Black Swans

As AI begins to mature, there are risks that lurk beneath the surface. When regulatory bodies begin to look at AI-assisted diagnostics or therapeutics as the new standard of care, reimbursement models may adjust, and providers may be left behind. The rapid proliferation and haphazard adoption of AI could lead to overdependence and deskilling or result in weird and as yet unknown errors that are difficult to troubleshoot.

What is the role of the GI societies? 

Specialty societies like AGA are taking leadership roles in determining the bounds of where AIs may tread, not just in providing information to their membership but also in digesting evidence and synthesizing recommendations. Societies must balance the real promise of AI in endoscopy with the practice realities for members, and provide living guidelines that reflect the consensus of members regarding scope of practice with the ability to update as new data become available.⁵

Societies also have a role as advocates for safety, taking ownership of high-quality content to prevent misinformation. AGA recently announced the development of a chat interface that will be focused on providing its members the highest quality information, and serve as a portal to identify and respond to its members’ information needs. By staying united rather than fragmenting, societies can maintain bounds to protect its members and their patients and advance areas where there is clinical need, together.

Dr. Shung is senior associate consultant, Division of Gastroenterology and Hepatology, and director of clinical generative artificial intelligence and informatics, Department of Medicine, at Mayo Clinic Rochester, Minnesota. He has no disclosures in regard to this article.

References

1. Soleymanjahi S, et al. Artificial Intelligence-Assisted Colonoscopy for Polyp Detection : A Systematic Review and Meta-analysis. Ann Intern Med. 2024 Dec. doi:10.7326/annals-24-00981.

2. Jong MR, et al. GastroNet-5M: A Multicenter Dataset for Developing Foundation Models in Gastrointestinal Endoscopy. Gastroenterology. 2025 Jul. doi: 10.1053/j.gastro.2025.07.030.

3. Soroush A, et al. Generative Artificial Intelligence in Clinical Medicine and Impact on Gastroenterology. Gastroenterology. 2025 Aug. doi: 10.1053/j.gastro.2025.03.038.

4. Mengying Fang C, et al. How AI and Human Behaviors Shape Psychosocial Effects of Extended Chatbot Use: A Longitudinal Randomized Controlled Study. arXiv e-prints. 2025 Mar. doi: 10.48550/arXiv.2503.17473.

5. Sultan S, et al. AGA Living Clinical Practice Guideline on Computer-Aided Detection-Assisted Colonoscopy. Gastroenterology. 2025 Apr. doi:10.1053/j.gastro.2025.01.002.

Dear colleagues,

Since our last Perspectives feature on artificial intelligence (AI) in gastroenterology and hepatology, the field has experienced remarkable growth in both innovation and clinical adoption. AI tools that were once conceptual are now entering everyday practice, with many more on the horizon poised to transform how we diagnose, treat, and manage patients. In this issue of Perspectives, we present two timely essays that explore how AI is reshaping clinical care—while also emphasizing the need for caution, thoughtful integration, and ongoing oversight.

Dr. Yuvaraj Singh, Dr. Alessandro Colletta, and Dr. Neil Marya discuss how purpose-built AI models can reduce diagnostic uncertainty in advanced endoscopy. From cholangioscopy systems that outperform standard ERCP sampling in distinguishing malignant biliary strictures to EUS-based platforms that differentiate autoimmune pancreatitis from pancreatic cancer, they envision a near-term future in which machine intelligence enhances accuracy, accelerates decision-making, and refines interpretation—without replacing the clinician’s expertise.

Complementing this, Dr. Dennis Shung takes a broader view across the endoscopy unit and outpatient clinic. He highlights the promise of AI for polyp detection, digital biopsy, and automated reporting, while underscoring the importance of human oversight, workflow integration, and safeguards against misinformation. Dr. Shung also emphasizes the pivotal role professional societies can play in establishing clear standards, ethical boundaries, and trusted frameworks for AI deployment in GI practice.

We hope these perspectives spark practical conversations about when—and how—to integrate AI in your own practice. As always, we welcome your feedback and real-world experience. Join the conversation on X at @AGA_GIHN.

Gyanprakash A. Ketwaroo, MD, MSc, is associate professor of medicine, Yale University, New Haven, and chief of endoscopy at West Haven VA Medical Center, both in Connecticut. He is an associate editor for GI & Hepatology News.

AI Models in Advanced Endoscopy

BY YUVARAJ SINGH, MD; ALESSANDRO COLLETTA, MD; NEIL MARYA, MD

As the adage goes, “if tumor is the rumor, then tissue is the issue, because cancer may be the answer.”

Establishing an accurate diagnosis is the essential first step toward curing or palliating malignancy. From detecting an early neoplastic lesion, to distinguishing between malignant and benign pathology, or to determining when and where to obtain tissue, endoscopists are frequently faced with the challenge of transforming diagnostic suspicion into certainty.

Artificial intelligence (AI), designed to replicate human cognition such as pattern recognition and decision-making, has emerged as a technology to assist gastroenterologists in addressing a variety of different tasks during endoscopy. AI research in gastrointestinal endoscopy has initially focused on computer-aided detection (CADe) of colorectal polyps. More recently, however, there has been increased emphasis on developing AI to assist advanced endoscopists.

For instance, in biliary endoscopy, AI is being explored to improve the notoriously challenging diagnosis of cholangiocarcinoma, where conventional tissue sampling often falls short of providing a definitive diagnosis. Similarly, in the pancreas, AI models are showing potential to differentiate autoimmune pancreatitis (AIP) from pancreatic ductal adenocarcinoma (PDAC), a distinction with profound therapeutic implications. Even pancreatic cysts are beginning to benefit from AI models that refine risk stratification and guide management. Together, these advances underscore how AI is not merely an adjunct but a potentially massive catalyst for reimagining the diagnostic role of advanced endoscopists.

Classifying biliary strictures (MBS) accurately remains a challenge. Standard ERCP-based sampling techniques (forceps biopsy and brush cytology) are suboptimal diagnostic tools with false negative rates for detecting MBS of less than 50%. The diagnostic uncertainty related to biliary stricture classification carries significant consequences for patients. For example, patients with biliary cancer without positive cytology have treatments delayed until a malignant diagnosis is established. 

Ancillary technologies to enhance ERCP-based tissue acquisition are still weighed down by low sensitivity and accuracy; even with ancillary use of fluorescent in situ hybridization (FISH), diagnostic yield remains limited. EUS-FNA can help with distal biliary strictures, but this technique risks needle-tract seeding in cases of perihilar disease. Cholangioscopy allows for direct visualization and targeted sampling; however, cholangioscopy-guided forceps biopsies are burdened by low sensitivities.¹ Additionally, physician interpretation of visual findings during cholangioscopy often suffers from poor interobserver agreement and poor accuracy.²

To improve the classification of biliary strictures, several groups have studied the application of AI for cholangioscopy footage of biliary pathology. In our lab, we trained an AI incorporating over 2.3 million cholangioscopy still images and nearly 20,000 expert-annotated frames to enhance its development. The AI closely mirrored expert labeling of cholangioscopy images of malignant pathology and, when tested on full cholangioscopy videos of indeterminate biliary strictures, the AI achieved a diagnostic accuracy of 91%—outperforming both brush cytology (63%) and forceps biopsy (61%).³

The results from this initial study were later validated across multiple centers. AI-assisted cholangioscopy could thus offer a reproducible, real-world solution to one of the most persistent diagnostic dilemmas advanced endoscopists face—helping clinicians act earlier and with greater confidence when evaluating indeterminate strictures.

Moving from the biliary tree to the pancreas, autoimmune pancreatitis (AIP) is a benign fibro-inflammatory disease that often frustrates advanced endoscopists as it closely mimics the appearance of pancreatic ductal adenocarcinoma (PDAC). The stakes are high: despite modern diagnostic techniques, including advanced imaging, some patients with pancreatic resections for “suspected PDAC” are still found to have AIP on final pathology. Conventional tools to distinguish AIP from PDAC have gaps: serum IgG4 and EUS-guided biopsies are both specific but insensitive.

Using EUS videos and images of various pancreas pathologies at Mayo Clinic, we developed an AI to tackle this dilemma. After intensive training, the EUS AI achieved a greater accuracy for distinguishing AIP from PDAC than a group of expert Mayo clinic endosonographers.⁵ In practice, an EUS-AI can identify AIP patterns in real-time, guiding clinicians toward steroid trials or biopsies and reducing the need for unnecessary surgeries.

Looking ahead, there are multiple opportunities for integration of AI into advanced endoscopy practices. Ongoing research suggests that AI could soon assist with identification of pancreas cysts most at risk for malignant transformation, classification of high risk Barrett’s esophagus, and even help with rapid on-site assessment of cytologic specimens obtained during EUS. Beyond diagnosis, AI could likely play an important role in guiding therapeutic interventions. For example, an ERCP AI in the future may be able to provide cannulation assistance or an AI assistant could help endosonographers during deployments of lumen apposing metal stents.

By enhancing image interpretation and procedural consistency, AI has the potential to uphold the fundamental principle of primum non nocere, enabling us to intervene with precision while minimizing harm. AI can also bridge grey zones in clinical practice and narrow diagnostic uncertainty in real time. Importantly, these systems can help clinicians achieve expertise in a fraction of the time it traditionally takes to acquire comparable human proficiency, while offering wider availability across practice settings and reducing interobserver variability that has long challenged endoscopic interpretation.

Currently, adoption is limited by high bias risk, lack of external validation, and interpretability Still, the trajectory of AI suggests a future where these computer technologies will not only support but also elevate human expertise, reshaping the standards of care of diseases managed by advanced endoscopists.

Dr. Singh, Dr. Colletta, and Dr. Marya are based at the Division of Gastroenterology and Hepatology, UMass Chan Medical School, Worcester, Massachusetts. Dr. Marya is a consultant for Boston Scientific, and has no other disclosures. Dr. Singh and Dr. Colletta have no disclosures.

References

1. Navaneethan U, et al. Comparative effectiveness of biliary brush cytology and intraductal biopsy for detection of malignant biliary strictures: a systematic review and meta-analysis. Gastrointest Endosc. 2015 Jan. doi: 10.1016/j.gie.2014.09.017.

2. Stassen PMC, et al. Diagnostic accuracy and interobserver agreement of digital single-operator cholangioscopy for indeterminate biliary strictures. Gastrointest Endosc 2021 Dec. doi: 10.1016/j.gie.2021.06.027.

3. Marya NB, et al. Identification of patients with malignant biliary strictures using a cholangioscopy-based deep learning artificial intelligence (with video). Gastrointest Endosc. 2023 Feb. doi: 10.1016/j.gie.2022.08.021.

4. Marya NB, et al. Multicenter validation of a cholangioscopy artificial intelligence system for the evaluation of biliary tract disease. Endoscopy. 2025 Aug. doi: 10.1055/a-2650-0789.

5. Marya NB, et al. Utilisation of artificial intelligence for the development of an EUS-convolutional neural network model trained to enhance the diagnosis of autoimmune pancreatitis. Gut. 2021 Jul. doi: 10.1136/gutjnl-2020-322821.

AI in General GI and Endoscopy

BY DENNIS L. SHUNG, MD, MHS, PHD

The practice of gastroenterology is changing, but much of it will be rooted in the same – careful, focused attention on endoscopic procedures, and compassionate, attentive care in clinic. Artificial intelligence (AI), like the Industrial Revolution before, is going to transform our practice. This comes with upsides and downsides, and highlights the need for strong leadership from our societies to safeguard the technology for practitioners and patients.

What are the upsides? 

AI has the potential to serve as a second set of eyes in detecting colon polyps, increasing the adenoma detection rate (ADR).¹ AI can be applied to all areas of the gastrointestinal tract, providing digital biopsies, guiding resection, and ensuring quality, which are all now possible with powerful new endoscopy foundation models, such as GastroNet-5M.²

Additionally. the advent of automating the collection of data into reports may herald the end of our days as data entry clerks. Generative AI also has the potential to give us all the best information at our fingertips, suggesting guideline-based care, providing the most up to date evidence, and guiding the differential diagnosis. The potential for patient-facing AI systems could lead to better health literacy, more meaningful engagement, and improved patient satisfaction.³

What are the downsides? 

For endoscopy, AI cannot make up for poor technique to ensure adequate mucosal exposure by the endoscopist, and an increase in AI-supported ADR does not yet convincingly translate into concrete gains in colorectal cancer-related mortality. For the foreseeable future, AI cannot make a connection with the patient in front of us, which is critical in diagnosing and treating patients.

Currently, AI appears to worsen loneliness⁴, and does not necessarily deepen the bonds or provide the positive touch that can heal, and which for many of us, was the reason we became physicians. Finally, as information proliferates, the information risk to patients and providers is growing – in the future, trusted sources to monitor, curate, and guide AI will be ever more important.

 

Black Swans

As AI begins to mature, there are risks that lurk beneath the surface. When regulatory bodies begin to look at AI-assisted diagnostics or therapeutics as the new standard of care, reimbursement models may adjust, and providers may be left behind. The rapid proliferation and haphazard adoption of AI could lead to overdependence and deskilling or result in weird and as yet unknown errors that are difficult to troubleshoot.

What is the role of the GI societies? 

Specialty societies like AGA are taking leadership roles in determining the bounds of where AIs may tread, not just in providing information to their membership but also in digesting evidence and synthesizing recommendations. Societies must balance the real promise of AI in endoscopy with the practice realities for members, and provide living guidelines that reflect the consensus of members regarding scope of practice with the ability to update as new data become available.⁵

Societies also have a role as advocates for safety, taking ownership of high-quality content to prevent misinformation. AGA recently announced the development of a chat interface that will be focused on providing its members the highest quality information, and serve as a portal to identify and respond to its members’ information needs. By staying united rather than fragmenting, societies can maintain bounds to protect its members and their patients and advance areas where there is clinical need, together.

Dr. Shung is senior associate consultant, Division of Gastroenterology and Hepatology, and director of clinical generative artificial intelligence and informatics, Department of Medicine, at Mayo Clinic Rochester, Minnesota. He has no disclosures in regard to this article.

References

1. Soleymanjahi S, et al. Artificial Intelligence-Assisted Colonoscopy for Polyp Detection : A Systematic Review and Meta-analysis. Ann Intern Med. 2024 Dec. doi:10.7326/annals-24-00981.

2. Jong MR, et al. GastroNet-5M: A Multicenter Dataset for Developing Foundation Models in Gastrointestinal Endoscopy. Gastroenterology. 2025 Jul. doi: 10.1053/j.gastro.2025.07.030.

3. Soroush A, et al. Generative Artificial Intelligence in Clinical Medicine and Impact on Gastroenterology. Gastroenterology. 2025 Aug. doi: 10.1053/j.gastro.2025.03.038.

4. Mengying Fang C, et al. How AI and Human Behaviors Shape Psychosocial Effects of Extended Chatbot Use: A Longitudinal Randomized Controlled Study. arXiv e-prints. 2025 Mar. doi: 10.48550/arXiv.2503.17473.

5. Sultan S, et al. AGA Living Clinical Practice Guideline on Computer-Aided Detection-Assisted Colonoscopy. Gastroenterology. 2025 Apr. doi:10.1053/j.gastro.2025.01.002.

Dear colleagues,

Since our last Perspectives feature on artificial intelligence (AI) in gastroenterology and hepatology, the field has experienced remarkable growth in both innovation and clinical adoption. AI tools that were once conceptual are now entering everyday practice, with many more on the horizon poised to transform how we diagnose, treat, and manage patients. In this issue of Perspectives, we present two timely essays that explore how AI is reshaping clinical care—while also emphasizing the need for caution, thoughtful integration, and ongoing oversight.

Dr. Yuvaraj Singh, Dr. Alessandro Colletta, and Dr. Neil Marya discuss how purpose-built AI models can reduce diagnostic uncertainty in advanced endoscopy. From cholangioscopy systems that outperform standard ERCP sampling in distinguishing malignant biliary strictures to EUS-based platforms that differentiate autoimmune pancreatitis from pancreatic cancer, they envision a near-term future in which machine intelligence enhances accuracy, accelerates decision-making, and refines interpretation—without replacing the clinician’s expertise.

Complementing this, Dr. Dennis Shung takes a broader view across the endoscopy unit and outpatient clinic. He highlights the promise of AI for polyp detection, digital biopsy, and automated reporting, while underscoring the importance of human oversight, workflow integration, and safeguards against misinformation. Dr. Shung also emphasizes the pivotal role professional societies can play in establishing clear standards, ethical boundaries, and trusted frameworks for AI deployment in GI practice.

We hope these perspectives spark practical conversations about when—and how—to integrate AI in your own practice. As always, we welcome your feedback and real-world experience. Join the conversation on X at @AGA_GIHN.

Gyanprakash A. Ketwaroo, MD, MSc, is associate professor of medicine, Yale University, New Haven, and chief of endoscopy at West Haven VA Medical Center, both in Connecticut. He is an associate editor for GI & Hepatology News.

AI Models in Advanced Endoscopy

BY YUVARAJ SINGH, MD; ALESSANDRO COLLETTA, MD; NEIL MARYA, MD

As the adage goes, “if tumor is the rumor, then tissue is the issue, because cancer may be the answer.”

Establishing an accurate diagnosis is the essential first step toward curing or palliating malignancy. From detecting an early neoplastic lesion, to distinguishing between malignant and benign pathology, or to determining when and where to obtain tissue, endoscopists are frequently faced with the challenge of transforming diagnostic suspicion into certainty.

Artificial intelligence (AI), designed to replicate human cognition such as pattern recognition and decision-making, has emerged as a technology to assist gastroenterologists in addressing a variety of different tasks during endoscopy. AI research in gastrointestinal endoscopy has initially focused on computer-aided detection (CADe) of colorectal polyps. More recently, however, there has been increased emphasis on developing AI to assist advanced endoscopists.

For instance, in biliary endoscopy, AI is being explored to improve the notoriously challenging diagnosis of cholangiocarcinoma, where conventional tissue sampling often falls short of providing a definitive diagnosis. Similarly, in the pancreas, AI models are showing potential to differentiate autoimmune pancreatitis (AIP) from pancreatic ductal adenocarcinoma (PDAC), a distinction with profound therapeutic implications. Even pancreatic cysts are beginning to benefit from AI models that refine risk stratification and guide management. Together, these advances underscore how AI is not merely an adjunct but a potentially massive catalyst for reimagining the diagnostic role of advanced endoscopists.

Classifying biliary strictures (MBS) accurately remains a challenge. Standard ERCP-based sampling techniques (forceps biopsy and brush cytology) are suboptimal diagnostic tools with false negative rates for detecting MBS of less than 50%. The diagnostic uncertainty related to biliary stricture classification carries significant consequences for patients. For example, patients with biliary cancer without positive cytology have treatments delayed until a malignant diagnosis is established. 

Ancillary technologies to enhance ERCP-based tissue acquisition are still weighed down by low sensitivity and accuracy; even with ancillary use of fluorescent in situ hybridization (FISH), diagnostic yield remains limited. EUS-FNA can help with distal biliary strictures, but this technique risks needle-tract seeding in cases of perihilar disease. Cholangioscopy allows for direct visualization and targeted sampling; however, cholangioscopy-guided forceps biopsies are burdened by low sensitivities.¹ Additionally, physician interpretation of visual findings during cholangioscopy often suffers from poor interobserver agreement and poor accuracy.²

To improve the classification of biliary strictures, several groups have studied the application of AI for cholangioscopy footage of biliary pathology. In our lab, we trained an AI incorporating over 2.3 million cholangioscopy still images and nearly 20,000 expert-annotated frames to enhance its development. The AI closely mirrored expert labeling of cholangioscopy images of malignant pathology and, when tested on full cholangioscopy videos of indeterminate biliary strictures, the AI achieved a diagnostic accuracy of 91%—outperforming both brush cytology (63%) and forceps biopsy (61%).³

The results from this initial study were later validated across multiple centers. AI-assisted cholangioscopy could thus offer a reproducible, real-world solution to one of the most persistent diagnostic dilemmas advanced endoscopists face—helping clinicians act earlier and with greater confidence when evaluating indeterminate strictures.

Moving from the biliary tree to the pancreas, autoimmune pancreatitis (AIP) is a benign fibro-inflammatory disease that often frustrates advanced endoscopists as it closely mimics the appearance of pancreatic ductal adenocarcinoma (PDAC). The stakes are high: despite modern diagnostic techniques, including advanced imaging, some patients with pancreatic resections for “suspected PDAC” are still found to have AIP on final pathology. Conventional tools to distinguish AIP from PDAC have gaps: serum IgG4 and EUS-guided biopsies are both specific but insensitive.

Using EUS videos and images of various pancreas pathologies at Mayo Clinic, we developed an AI to tackle this dilemma. After intensive training, the EUS AI achieved a greater accuracy for distinguishing AIP from PDAC than a group of expert Mayo clinic endosonographers.⁵ In practice, an EUS-AI can identify AIP patterns in real-time, guiding clinicians toward steroid trials or biopsies and reducing the need for unnecessary surgeries.

Looking ahead, there are multiple opportunities for integration of AI into advanced endoscopy practices. Ongoing research suggests that AI could soon assist with identification of pancreas cysts most at risk for malignant transformation, classification of high risk Barrett’s esophagus, and even help with rapid on-site assessment of cytologic specimens obtained during EUS. Beyond diagnosis, AI could likely play an important role in guiding therapeutic interventions. For example, an ERCP AI in the future may be able to provide cannulation assistance or an AI assistant could help endosonographers during deployments of lumen apposing metal stents.

By enhancing image interpretation and procedural consistency, AI has the potential to uphold the fundamental principle of primum non nocere, enabling us to intervene with precision while minimizing harm. AI can also bridge grey zones in clinical practice and narrow diagnostic uncertainty in real time. Importantly, these systems can help clinicians achieve expertise in a fraction of the time it traditionally takes to acquire comparable human proficiency, while offering wider availability across practice settings and reducing interobserver variability that has long challenged endoscopic interpretation.

Currently, adoption is limited by high bias risk, lack of external validation, and interpretability Still, the trajectory of AI suggests a future where these computer technologies will not only support but also elevate human expertise, reshaping the standards of care of diseases managed by advanced endoscopists.

Dr. Singh, Dr. Colletta, and Dr. Marya are based at the Division of Gastroenterology and Hepatology, UMass Chan Medical School, Worcester, Massachusetts. Dr. Marya is a consultant for Boston Scientific, and has no other disclosures. Dr. Singh and Dr. Colletta have no disclosures.

References

1. Navaneethan U, et al. Comparative effectiveness of biliary brush cytology and intraductal biopsy for detection of malignant biliary strictures: a systematic review and meta-analysis. Gastrointest Endosc. 2015 Jan. doi: 10.1016/j.gie.2014.09.017.

2. Stassen PMC, et al. Diagnostic accuracy and interobserver agreement of digital single-operator cholangioscopy for indeterminate biliary strictures. Gastrointest Endosc 2021 Dec. doi: 10.1016/j.gie.2021.06.027.

3. Marya NB, et al. Identification of patients with malignant biliary strictures using a cholangioscopy-based deep learning artificial intelligence (with video). Gastrointest Endosc. 2023 Feb. doi: 10.1016/j.gie.2022.08.021.

4. Marya NB, et al. Multicenter validation of a cholangioscopy artificial intelligence system for the evaluation of biliary tract disease. Endoscopy. 2025 Aug. doi: 10.1055/a-2650-0789.

5. Marya NB, et al. Utilisation of artificial intelligence for the development of an EUS-convolutional neural network model trained to enhance the diagnosis of autoimmune pancreatitis. Gut. 2021 Jul. doi: 10.1136/gutjnl-2020-322821.

AI in General GI and Endoscopy

BY DENNIS L. SHUNG, MD, MHS, PHD

The practice of gastroenterology is changing, but much of it will be rooted in the same – careful, focused attention on endoscopic procedures, and compassionate, attentive care in clinic. Artificial intelligence (AI), like the Industrial Revolution before, is going to transform our practice. This comes with upsides and downsides, and highlights the need for strong leadership from our societies to safeguard the technology for practitioners and patients.

What are the upsides? 

AI has the potential to serve as a second set of eyes in detecting colon polyps, increasing the adenoma detection rate (ADR).¹ AI can be applied to all areas of the gastrointestinal tract, providing digital biopsies, guiding resection, and ensuring quality, which are all now possible with powerful new endoscopy foundation models, such as GastroNet-5M.²

Additionally. the advent of automating the collection of data into reports may herald the end of our days as data entry clerks. Generative AI also has the potential to give us all the best information at our fingertips, suggesting guideline-based care, providing the most up to date evidence, and guiding the differential diagnosis. The potential for patient-facing AI systems could lead to better health literacy, more meaningful engagement, and improved patient satisfaction.³

What are the downsides? 

For endoscopy, AI cannot make up for poor technique to ensure adequate mucosal exposure by the endoscopist, and an increase in AI-supported ADR does not yet convincingly translate into concrete gains in colorectal cancer-related mortality. For the foreseeable future, AI cannot make a connection with the patient in front of us, which is critical in diagnosing and treating patients.

Currently, AI appears to worsen loneliness⁴, and does not necessarily deepen the bonds or provide the positive touch that can heal, and which for many of us, was the reason we became physicians. Finally, as information proliferates, the information risk to patients and providers is growing – in the future, trusted sources to monitor, curate, and guide AI will be ever more important.

 

Black Swans

As AI begins to mature, there are risks that lurk beneath the surface. When regulatory bodies begin to look at AI-assisted diagnostics or therapeutics as the new standard of care, reimbursement models may adjust, and providers may be left behind. The rapid proliferation and haphazard adoption of AI could lead to overdependence and deskilling or result in weird and as yet unknown errors that are difficult to troubleshoot.

What is the role of the GI societies? 

Specialty societies like AGA are taking leadership roles in determining the bounds of where AIs may tread, not just in providing information to their membership but also in digesting evidence and synthesizing recommendations. Societies must balance the real promise of AI in endoscopy with the practice realities for members, and provide living guidelines that reflect the consensus of members regarding scope of practice with the ability to update as new data become available.⁵

Societies also have a role as advocates for safety, taking ownership of high-quality content to prevent misinformation. AGA recently announced the development of a chat interface that will be focused on providing its members the highest quality information, and serve as a portal to identify and respond to its members’ information needs. By staying united rather than fragmenting, societies can maintain bounds to protect its members and their patients and advance areas where there is clinical need, together.

Dr. Shung is senior associate consultant, Division of Gastroenterology and Hepatology, and director of clinical generative artificial intelligence and informatics, Department of Medicine, at Mayo Clinic Rochester, Minnesota. He has no disclosures in regard to this article.

References

1. Soleymanjahi S, et al. Artificial Intelligence-Assisted Colonoscopy for Polyp Detection : A Systematic Review and Meta-analysis. Ann Intern Med. 2024 Dec. doi:10.7326/annals-24-00981.

2. Jong MR, et al. GastroNet-5M: A Multicenter Dataset for Developing Foundation Models in Gastrointestinal Endoscopy. Gastroenterology. 2025 Jul. doi: 10.1053/j.gastro.2025.07.030.

3. Soroush A, et al. Generative Artificial Intelligence in Clinical Medicine and Impact on Gastroenterology. Gastroenterology. 2025 Aug. doi: 10.1053/j.gastro.2025.03.038.

4. Mengying Fang C, et al. How AI and Human Behaviors Shape Psychosocial Effects of Extended Chatbot Use: A Longitudinal Randomized Controlled Study. arXiv e-prints. 2025 Mar. doi: 10.48550/arXiv.2503.17473.

5. Sultan S, et al. AGA Living Clinical Practice Guideline on Computer-Aided Detection-Assisted Colonoscopy. Gastroenterology. 2025 Apr. doi:10.1053/j.gastro.2025.01.002.

The requisites of government are that there be sufficiency of food, sufficiency of military equipment, and the confidence of the people in their ruler.

Analects by Confucius¹

From ancient festivals to modern holidays, autumn has long been associated with the gathering of the harvest. Friends and families come together around tables laden with delicious food to enjoy the pleasures of peace and plenty. During these celebrations, we must never forget that without the strength of the nation’s military and the service of its veterans, this freedom and abundance would not be possible. Our debt of gratitude to the current and former members of the armed services makes the fact that a substantial minority experiences food insecurity not only a human tragedy, but a travesty of the nation’s promise to support those who wear or have worn the uniform.

The National Defense Authorization Act for Fiscal Year 2020 charged the Secretary of Defense to investigate food insecurity among active-duty service members and their dependents.² The RAND Corporation conducted the assessment and, based on the results of its analysis, made recommendations to reduce hunger among armed forces members and their families.³

The RAND study found that 10% of active-duty military met US Department of Agriculture (USDA) criteria for very low food security; another 15% were classified as having low food security. The USDA defines food insecurity with hunger as “reports of multiple indications of disrupted eating patterns and reduced food intake.” USDA defines low food security as “reports of reduced quality, variety, or desirability of diet. Little or no indication of reduced food intake.”⁴

As someone who grew up on an Army base with the commissary a short trip from military housing, I was unpleasantly surprised that food insecurity was more common among in-service members living on post. I was even more dismayed to read that a variety of factors constrained 14% of active-duty military experiencing food insecurity to seek public assistance to feed themselves and their families. As with so many health care and social services, (eg, mental health care), those wearing the uniform were concerned that participating in a food assistance program would damage their career or stigmatize them. Others did not seek help, perhaps because they believed they were not eligible, and in many cases were correct: they did not qualify for food banks or food stamps due to receiving other benefits. A variety of factors contribute to periods of food insecurity among military families, including remote or rural bases that lack access to grocery stores or jobs for partners or other family members, and low base military pay.⁵

Food insecurity is an even more serious concern among veterans who are frequently older and have more comorbidities, often leading to unemployment and homelessness. Feeding America, the nation’s largest organization of community food banks, estimates that 1 in 9 working-age veterans are food insecure.⁵ US Department of Veterans Affairs (VA) statistics indicate that veterans are 7% more likely to experience food insecurity than other sectors of the population.⁶ The Veterans Health Administration has recognized that food insecurity is directly related to medical problems already common among veterans, including diabetes, obesity, and depression. Women and minority veterans are the most at risk of food insecurity.⁷

Recognizing that many veterans are at risk of food insecurity, the US Department of Defense and VA have taken steps to try and reduce hunger among those who serve. In response to the shocking statistic that food insecurity was found in 27% of Iraq and Afghanistan veterans, the VA and Rockefeller Foundation are partnering on the Food as Medicine initiative to improve veteran nutrition as a means of improving nutrition-related health consequences of food insecurity.⁸

Like many federal practitioners, I was unaware of the food insecurity assistance available to active-duty service members or veterans, or how to help individuals access it. In addition to the resources outlined in the Table, there are many community-based options open to anyone, including veterans and service members. 

I have written columns on many difficult issues in my years as the Editor-in-Chief of Federal Practitioner, but personally this is one of the most distressing editorials I have ever published. That individuals dedicated to defending our rights and protecting our safety should be compelled to go hungry or not know if they have enough money at the end of the month to buy food is manifestly unjust. It is challenging when faced with such a large-scale injustice to think we cannot make a difference, but that resignation or abdication only magnifies this inequity. I have a friend who kept giving back even after they retired from federal service: they volunteered at a community garden and brought produce to the local food bank and helped distribute it. That may seem too much for those still working yet almost anyone can pick up a few items on their weekly shopping trip and donate them to a food drive. 

As we approach Veterans Day, let’s not just express our gratitude to our military and veterans in words but in deeds like feeding the hungry and urging elected representatives to fulfill their commitment to ensure that service members and veterans and their families do not experience food insecurity. Confucian wisdom written in a very distant time and vastly dissimilar context still rings true: there are direct and critical links between food and trust and between hunger and the military.¹

The requisites of government are that there be sufficiency of food, sufficiency of military equipment, and the confidence of the people in their ruler.

Analects by Confucius¹

From ancient festivals to modern holidays, autumn has long been associated with the gathering of the harvest. Friends and families come together around tables laden with delicious food to enjoy the pleasures of peace and plenty. During these celebrations, we must never forget that without the strength of the nation’s military and the service of its veterans, this freedom and abundance would not be possible. Our debt of gratitude to the current and former members of the armed services makes the fact that a substantial minority experiences food insecurity not only a human tragedy, but a travesty of the nation’s promise to support those who wear or have worn the uniform.

The National Defense Authorization Act for Fiscal Year 2020 charged the Secretary of Defense to investigate food insecurity among active-duty service members and their dependents.² The RAND Corporation conducted the assessment and, based on the results of its analysis, made recommendations to reduce hunger among armed forces members and their families.³

The RAND study found that 10% of active-duty military met US Department of Agriculture (USDA) criteria for very low food security; another 15% were classified as having low food security. The USDA defines food insecurity with hunger as “reports of multiple indications of disrupted eating patterns and reduced food intake.” USDA defines low food security as “reports of reduced quality, variety, or desirability of diet. Little or no indication of reduced food intake.”⁴

As someone who grew up on an Army base with the commissary a short trip from military housing, I was unpleasantly surprised that food insecurity was more common among in-service members living on post. I was even more dismayed to read that a variety of factors constrained 14% of active-duty military experiencing food insecurity to seek public assistance to feed themselves and their families. As with so many health care and social services, (eg, mental health care), those wearing the uniform were concerned that participating in a food assistance program would damage their career or stigmatize them. Others did not seek help, perhaps because they believed they were not eligible, and in many cases were correct: they did not qualify for food banks or food stamps due to receiving other benefits. A variety of factors contribute to periods of food insecurity among military families, including remote or rural bases that lack access to grocery stores or jobs for partners or other family members, and low base military pay.⁵

Food insecurity is an even more serious concern among veterans who are frequently older and have more comorbidities, often leading to unemployment and homelessness. Feeding America, the nation’s largest organization of community food banks, estimates that 1 in 9 working-age veterans are food insecure.⁵ US Department of Veterans Affairs (VA) statistics indicate that veterans are 7% more likely to experience food insecurity than other sectors of the population.⁶ The Veterans Health Administration has recognized that food insecurity is directly related to medical problems already common among veterans, including diabetes, obesity, and depression. Women and minority veterans are the most at risk of food insecurity.⁷

Recognizing that many veterans are at risk of food insecurity, the US Department of Defense and VA have taken steps to try and reduce hunger among those who serve. In response to the shocking statistic that food insecurity was found in 27% of Iraq and Afghanistan veterans, the VA and Rockefeller Foundation are partnering on the Food as Medicine initiative to improve veteran nutrition as a means of improving nutrition-related health consequences of food insecurity.⁸

Like many federal practitioners, I was unaware of the food insecurity assistance available to active-duty service members or veterans, or how to help individuals access it. In addition to the resources outlined in the Table, there are many community-based options open to anyone, including veterans and service members. 

I have written columns on many difficult issues in my years as the Editor-in-Chief of Federal Practitioner, but personally this is one of the most distressing editorials I have ever published. That individuals dedicated to defending our rights and protecting our safety should be compelled to go hungry or not know if they have enough money at the end of the month to buy food is manifestly unjust. It is challenging when faced with such a large-scale injustice to think we cannot make a difference, but that resignation or abdication only magnifies this inequity. I have a friend who kept giving back even after they retired from federal service: they volunteered at a community garden and brought produce to the local food bank and helped distribute it. That may seem too much for those still working yet almost anyone can pick up a few items on their weekly shopping trip and donate them to a food drive. 

As we approach Veterans Day, let’s not just express our gratitude to our military and veterans in words but in deeds like feeding the hungry and urging elected representatives to fulfill their commitment to ensure that service members and veterans and their families do not experience food insecurity. Confucian wisdom written in a very distant time and vastly dissimilar context still rings true: there are direct and critical links between food and trust and between hunger and the military.¹

The requisites of government are that there be sufficiency of food, sufficiency of military equipment, and the confidence of the people in their ruler.

Analects by Confucius¹

From ancient festivals to modern holidays, autumn has long been associated with the gathering of the harvest. Friends and families come together around tables laden with delicious food to enjoy the pleasures of peace and plenty. During these celebrations, we must never forget that without the strength of the nation’s military and the service of its veterans, this freedom and abundance would not be possible. Our debt of gratitude to the current and former members of the armed services makes the fact that a substantial minority experiences food insecurity not only a human tragedy, but a travesty of the nation’s promise to support those who wear or have worn the uniform.

The National Defense Authorization Act for Fiscal Year 2020 charged the Secretary of Defense to investigate food insecurity among active-duty service members and their dependents.² The RAND Corporation conducted the assessment and, based on the results of its analysis, made recommendations to reduce hunger among armed forces members and their families.³

The RAND study found that 10% of active-duty military met US Department of Agriculture (USDA) criteria for very low food security; another 15% were classified as having low food security. The USDA defines food insecurity with hunger as “reports of multiple indications of disrupted eating patterns and reduced food intake.” USDA defines low food security as “reports of reduced quality, variety, or desirability of diet. Little or no indication of reduced food intake.”⁴

As someone who grew up on an Army base with the commissary a short trip from military housing, I was unpleasantly surprised that food insecurity was more common among in-service members living on post. I was even more dismayed to read that a variety of factors constrained 14% of active-duty military experiencing food insecurity to seek public assistance to feed themselves and their families. As with so many health care and social services, (eg, mental health care), those wearing the uniform were concerned that participating in a food assistance program would damage their career or stigmatize them. Others did not seek help, perhaps because they believed they were not eligible, and in many cases were correct: they did not qualify for food banks or food stamps due to receiving other benefits. A variety of factors contribute to periods of food insecurity among military families, including remote or rural bases that lack access to grocery stores or jobs for partners or other family members, and low base military pay.⁵

Food insecurity is an even more serious concern among veterans who are frequently older and have more comorbidities, often leading to unemployment and homelessness. Feeding America, the nation’s largest organization of community food banks, estimates that 1 in 9 working-age veterans are food insecure.⁵ US Department of Veterans Affairs (VA) statistics indicate that veterans are 7% more likely to experience food insecurity than other sectors of the population.⁶ The Veterans Health Administration has recognized that food insecurity is directly related to medical problems already common among veterans, including diabetes, obesity, and depression. Women and minority veterans are the most at risk of food insecurity.⁷

Recognizing that many veterans are at risk of food insecurity, the US Department of Defense and VA have taken steps to try and reduce hunger among those who serve. In response to the shocking statistic that food insecurity was found in 27% of Iraq and Afghanistan veterans, the VA and Rockefeller Foundation are partnering on the Food as Medicine initiative to improve veteran nutrition as a means of improving nutrition-related health consequences of food insecurity.⁸

Like many federal practitioners, I was unaware of the food insecurity assistance available to active-duty service members or veterans, or how to help individuals access it. In addition to the resources outlined in the Table, there are many community-based options open to anyone, including veterans and service members. 

I have written columns on many difficult issues in my years as the Editor-in-Chief of Federal Practitioner, but personally this is one of the most distressing editorials I have ever published. That individuals dedicated to defending our rights and protecting our safety should be compelled to go hungry or not know if they have enough money at the end of the month to buy food is manifestly unjust. It is challenging when faced with such a large-scale injustice to think we cannot make a difference, but that resignation or abdication only magnifies this inequity. I have a friend who kept giving back even after they retired from federal service: they volunteered at a community garden and brought produce to the local food bank and helped distribute it. That may seem too much for those still working yet almost anyone can pick up a few items on their weekly shopping trip and donate them to a food drive. 

As we approach Veterans Day, let’s not just express our gratitude to our military and veterans in words but in deeds like feeding the hungry and urging elected representatives to fulfill their commitment to ensure that service members and veterans and their families do not experience food insecurity. Confucian wisdom written in a very distant time and vastly dissimilar context still rings true: there are direct and critical links between food and trust and between hunger and the military.¹

An October 1 article in GI & Hepatology News cautioned physicians against partnering with private equity firms, warning that they target “quick profits and quick exits, which can be inconsistent with quality long-term patient care.”

But several recent studies – and my own experience – show that affiliating with a private equity-backed management services organization can empower physician practices to deliver high-quality care at lower cost than other practice affiliation models.

A 2024 study conducted by Avalere Health found that per-beneficiary Medicare expenditures for physicians who shifted from an unaffiliated practice model to a PE-affiliated model declined by $963 in the 12 months following the transition. By contrast, per-beneficiary Medicare expenditures for physicians who shifted from an unaffiliated model to a hospital-affiliated one increased more than $1,300.

A 2025 peer-reviewed study published in Journal of Market Access & Health Policy found that physicians affiliated with private equity were far more likely to perform common high-volume procedures in the lowest-cost site of care – an ambulatory surgery center or medical office – than in higher-cost hospital outpatient departments. Physicians affiliated with hospitals were far more likely to perform procedures in HOPDs.

Partnering with a private equity-backed management services organization has enabled my practice to afford advanced technologies we never could have deployed on our own. Those technologies have helped improve our polyp detection rates, reduce the incidence of colon cancer, and more efficiently care for patients with ulcerative colitis. We also now provide patients seamless access to digital platforms that help them better manage chronic conditions.

Independent medical practice is under duress. Partnering with a private equity-backed management services organization is one of the most effective ways for a physician practice to retain its independence – and continue offering patients affordable, high-quality care.

George Dickstein, MD, AGAF, is senior vice president of clinical affairs, Massachusetts, for Gastro Health, and chairperson of Gastro Health’s Physician Leadership Council. He is based in Framingham, Mass. GI & Hepatology News encourages readers to submit letters to the editor to debate topics raised in the newspaper.

An October 1 article in GI & Hepatology News cautioned physicians against partnering with private equity firms, warning that they target “quick profits and quick exits, which can be inconsistent with quality long-term patient care.”

But several recent studies – and my own experience – show that affiliating with a private equity-backed management services organization can empower physician practices to deliver high-quality care at lower cost than other practice affiliation models.

A 2024 study conducted by Avalere Health found that per-beneficiary Medicare expenditures for physicians who shifted from an unaffiliated practice model to a PE-affiliated model declined by $963 in the 12 months following the transition. By contrast, per-beneficiary Medicare expenditures for physicians who shifted from an unaffiliated model to a hospital-affiliated one increased more than $1,300.

A 2025 peer-reviewed study published in Journal of Market Access & Health Policy found that physicians affiliated with private equity were far more likely to perform common high-volume procedures in the lowest-cost site of care – an ambulatory surgery center or medical office – than in higher-cost hospital outpatient departments. Physicians affiliated with hospitals were far more likely to perform procedures in HOPDs.

Partnering with a private equity-backed management services organization has enabled my practice to afford advanced technologies we never could have deployed on our own. Those technologies have helped improve our polyp detection rates, reduce the incidence of colon cancer, and more efficiently care for patients with ulcerative colitis. We also now provide patients seamless access to digital platforms that help them better manage chronic conditions.

Independent medical practice is under duress. Partnering with a private equity-backed management services organization is one of the most effective ways for a physician practice to retain its independence – and continue offering patients affordable, high-quality care.

George Dickstein, MD, AGAF, is senior vice president of clinical affairs, Massachusetts, for Gastro Health, and chairperson of Gastro Health’s Physician Leadership Council. He is based in Framingham, Mass. GI & Hepatology News encourages readers to submit letters to the editor to debate topics raised in the newspaper.

An October 1 article in GI & Hepatology News cautioned physicians against partnering with private equity firms, warning that they target “quick profits and quick exits, which can be inconsistent with quality long-term patient care.”

But several recent studies – and my own experience – show that affiliating with a private equity-backed management services organization can empower physician practices to deliver high-quality care at lower cost than other practice affiliation models.

A 2024 study conducted by Avalere Health found that per-beneficiary Medicare expenditures for physicians who shifted from an unaffiliated practice model to a PE-affiliated model declined by $963 in the 12 months following the transition. By contrast, per-beneficiary Medicare expenditures for physicians who shifted from an unaffiliated model to a hospital-affiliated one increased more than $1,300.

A 2025 peer-reviewed study published in Journal of Market Access & Health Policy found that physicians affiliated with private equity were far more likely to perform common high-volume procedures in the lowest-cost site of care – an ambulatory surgery center or medical office – than in higher-cost hospital outpatient departments. Physicians affiliated with hospitals were far more likely to perform procedures in HOPDs.

Partnering with a private equity-backed management services organization has enabled my practice to afford advanced technologies we never could have deployed on our own. Those technologies have helped improve our polyp detection rates, reduce the incidence of colon cancer, and more efficiently care for patients with ulcerative colitis. We also now provide patients seamless access to digital platforms that help them better manage chronic conditions.

Independent medical practice is under duress. Partnering with a private equity-backed management services organization is one of the most effective ways for a physician practice to retain its independence – and continue offering patients affordable, high-quality care.

George Dickstein, MD, AGAF, is senior vice president of clinical affairs, Massachusetts, for Gastro Health, and chairperson of Gastro Health’s Physician Leadership Council. He is based in Framingham, Mass. GI & Hepatology News encourages readers to submit letters to the editor to debate topics raised in the newspaper.