Hidradenitis suppurativa (HS) is a chronic inflammatory, androgen gland disorder characterized by recurrent rupture of the hair follicles with a vigorous inflammatory response. This response results in abscess formation and development of draining sinus tracts and hypertrophic fibrous scars.^1,2 Pain, discomfort, and odorous discharge from the recalcitrant lesions have a profound impact on patient quality of life.^3,4

The morbidity and disease burden associated with HS are particularly underestimated, as patients frequently report debilitating pain that often is overlooked.^5,6 Additionally, the quality and intensity of perceived pain are compounded by frequently associated depression and anxiety.^7-9 Pain has been reported by patients with HS to be the highest cause of morbidity, despite the disfiguring nature of the disease and its associated psychosocial distress.^7,10 Nonetheless, HS lacks an accepted pain management algorithm similar to those that have been developed for the treatment of other acute or chronic pain disorders, such as back pain and sickle cell disease.^4,11-13

Given the lack of formal studies regarding pain management in patients with HS, clinicians are limited to general pain guidelines, expert opinion, small trials, and patient preference.³ Furthermore, effective pain management in HS necessitates the treatment of both chronic pain affecting daily function and acute pain present during disease flares, surgical interventions, and dressing changes.³ The result is a wide array of strategies used for HS-associated pain.^3,4

Epidemiology and Pathophysiology

Hidradenitis suppurativa historically has been an overlooked and underdiagnosed disease, which limits epidemiology data.⁵ Current estimates are that HS affects approximately 1% of the general population; however, prevalence rates range from 0.03% to 4.1%.^14-16

The exact etiology of HS remains unclear, but it is thought that genetic factors, immune dysregulation, and environmental/behavioral influences all contribute to its pathophysiology.^1,17 Up to 40% of patients with HS report a positive family history of the disease.^18-20 Hidradenitis suppurativa has been associated with other inflammatory disease states, such as inflammatory bowel disease, spondyloarthropathies, and pyoderma gangrenosum.^16,21,22

It is thought that HS is the result of some defect in keratin clearance that leads to follicular hyperkeratinization and occlusion.¹ Resultant rupture of pilosebaceous units and spillage of contents (including keratin and bacteria) into the surrounding dermis triggers a vigorous inflammatory response. Sinus tracts and fistulas become the targets of bacterial colonization, biofilm formation, and secondary infection. The result is suppuration and extension of the lesions as well as sustained chronic inflammation.^23,24

Although the etiology of HS is complex, several modifiable risk factors for the disease have been identified, most prominently cigarette smoking and obesity. Approximately 70% of patients with HS smoke cigarettes.^2,15,25,26 Obesity has a well-known association with HS, and it is possible that weight reduction lowers disease severity.^27-30

Clinical Presentation and Diagnosis

Establishing a diagnosis of HS necessitates recognition of disease morphology, topography, and chronicity. Hidradenitis suppurativa most commonly occurs in the axillae, inguinal and anogenital region, perineal region, and inframammary region.^5,31 A typical history involves a prolonged disease course with recurrent lesions and intermittent periods of improvement or remission. Primary lesions are deep, inflamed, painful, and sterile. Ultimately, these lesions rupture and track subcutaneously.^15,25 Intercommunicating sinus tracts form from multiple recurrent nodules in close proximity and may ultimately lead to fibrotic scarring and local architectural distortion.³² The Hurley staging system helps to guide treatment interventions based on disease severity. Approach to pain management is discussed below.

Pain Management in HS: General Principles

Pain management is complex for clinicians, as there are limited studies from which to draw treatment recommendations. Incomplete understanding of the etiology and pathophysiology of the disease contributes to the lack of established management guidelines.

A PubMed search of articles indexed for MEDLINE using the terms hidradenitis, suppurativa, pain, and management revealed 61 different results dating back to 1980, 52 of which had been published in the last 5 years. When the word acute was added to the search, there were only 6 results identified. These results clearly reflect a better understanding of HS-mediated pain as well as clinical unmet needs and evolving strategies in pain management therapeutics. However, many of these studies reflect therapies focused on the mediation or modulation of HS pathogenesis rather than potential pain management therapies.

In addition, the heterogenous nature of the pain experience in HS poses a challenge for clinicians. Patients may experience multiple pain types concurrently, including inflammatory, noninflammatory, nociceptive, neuropathic, and ischemic, as well as pain related to arthritis.^3,33,34 Pain perception is further complicated by the observation that patients with HS have high rates of psychiatric comorbidities such as depression and anxiety, both of which profoundly alter perception of both the strength and quality of pain.^7,8,22,35 A suggested algorithm for treatment of pain in HS is described in the eTable.³⁶

Chronicity is a hallmark of HS. Patients experience a prolonged disease course involving acute painful exacerbations superimposed on chronic pain that affects all aspects of daily life. Changes in self-perception, daily living activities, mood state, physical functioning, and physical comfort frequently are reported to have a major impact on quality of life.^1,3,37

In 2018, Thorlacius et al³⁸ created a multistakeholder consensus on a core outcome set of domains detailing what to measure in clinical trials for HS. The authors hoped that the routine adoption of these core domains would promote the collection of consistent and relevant information, bolster the strength of evidence synthesis, and minimize the risk for outcome reporting bias among studies.³⁸ It is important to ascertain the patient’s description of his/her pain to distinguish between stimulus-dependent nociceptive pain vs spontaneous neuropathic pain.^3,7,10 The most common pain descriptors used by patients are “shooting,” “itchy,” “blinding,” “cutting,” and “exhausting.”¹⁰ In addition to obtaining descriptive factors, it is important for the clinician to obtain information on the timing of the pain, whether or not the pain is relieved with spontaneous or surgical drainage, and if the patient is experiencing chronic background pain secondary to scarring or skin contraction.³ With the routine utilization of a consistent set of core domains, advances in our understanding of the different elements of HS pain, and increased provider awareness of the disease, the future of pain management in patients with HS seems promising.

Acute and Perioperative Pain Management

Acute Pain Management—The pain in HS can range from mild to excruciating.^3,7 The difference between acute and chronic pain in this condition may be hard to delineate, as patients may have intense acute flares on top of a baseline level of chronic pain.^3,7,14 These factors, in combination with various pain types of differing etiologies, make the treatment of HS-associated pain a therapeutic challenge.

The first-line treatments for acute pain in HS are oral acetaminophen, oral nonsteroidal anti-inflammatory drugs (NSAIDs), and topical analgesics.³ These treatment modalities are especially helpful for nociceptive pain, which often is described as having an aching or tender quality.³ Topical treatment for acute pain episodes includes diclofenac gel and liposomal lidocaine cream.³⁹ Topical lidocaine in particular has the benefit of being rapid acting, and its effect can last 1 to 2 hours. Ketamine has been anecdotally used as a topical treatment. Treatment options for neuropathic pain include topical amitriptyline, gabapentin, and pregabalin.³⁹ Dressings and ice packs may be used in cases of mild acute pain, depending on patient preference.³

First-line therapies may not provide adequate pain control in many patients.^3,40,41 Should the first-line treatments fail, oral opiates can be considered as a treatment option, especially if the patient has a history of recurrent pain unresponsive to milder methods of pain control.^3,40,41 However, prudence should be exercised, as patients with HS have a higher risk for opioid abuse, and referral to a pain specialist is advisable.⁴⁰ Generally, use of opioids should be limited to the smallest period of time possible.^40,41 Codeine can be used as a first opioid option, with hydromorphone available as an alternative.⁴¹

Pain caused by inflamed abscesses and nodules can be treated with either intralesional corticosteroids or incision and drainage. Intralesional triamcinolone has been found to cause substantial pain relief within 1 day of injection in patients with HS.^3,42

Prompt discussion about the remitting course of HS will prepare patients for flares. Although the therapies discussed here aim to reduce the clinical severity and inflammation associated with HS, achieving pain-free remission can be challenging. Barriers to developing a long-term treatment regimen include intolerable side effects or simply nonresponsive disease.^36,43

Management of Perioperative Pain—Medical treatment of HS often yields only transient or mild results. Hurley stage II or III lesions typically require surgical removal of affected tissues.^32,44-46 Surgery may dramatically reduce the primary disease burden and provide substantial pain relief.^3,4,44 Complete resection of the affected tissue by wide excision is the most common surgical procedure used.^46-48 However, various tissue-sparing techniques, such as skin-tissue-sparing excision with electrosurgical peeling, also have been utilized. Tissue-sparing surgical techniques may lead to shorter healing times and less postoperative pain.⁴⁸

There currently is little guidance available on the perioperative management of pain as it relates to surgical procedures for HS. The pain experienced from surgery varies based on the area and location of affected tissue; extent of disease; surgical technique used; and whether primary closure, closure by secondary intention, or skin grafting is utilized.^47,49 Medical treatment aimed at reducing inflammation prior to surgical intervention may improve postoperative pain and complications.

The use of general vs local anesthesia during surgery depends on the extent of the disease and the amount of tissue being removed; however, the use of local anesthesia has been associated with a higher recurrence of disease, possibly owing to less aggressive tissue removal.⁵⁰ Intraoperatively, the injection of 0.5% bupivacaine around the wound edges may lead to less postoperative pain.^3,48 Postoperative pain usually is managed with acetaminophen and NSAIDs.⁴⁸ In cases of severe postoperative pain, short- and long-acting opioid oxycodone preparations may be used. The combination of diclofenac and tramadol also has been used postoperatively.³ Patients who do not undergo extensive surgery often can leave the hospital the same day.

Effective strategies for mitigating HS-associated pain must address the chronic pain component of the disease. Long-term management involves lifestyle modifications and pharmacologic agents.

Chronic Pain Management

Although HS is not a curable disease, there are treatments available to minimize symptoms. Long-term management of HS is essential to minimize the effects of chronic pain and physical scarring associated with inflammation.³¹ In one study from the French Society of Dermatology, pain reported by patients with HS was directly associated with severity and duration of disease, emotional symptoms, and reduced functionality.⁵¹ For these reasons, many treatments for HS target reducing clinical severity and achieving remission, often defined as more than 6 months without any recurrence of lesions.⁵² In addition to lifestyle management, therapies available to manage HS include topical and systemic medications as well as procedures such as surgical excision.^36,43,52,53

Lifestyle Modifications

Regardless of the severity of HS, all patients may benefit from basic education on the pathogenesis of the disease.³⁶ The associations with smoking and obesity have been well documented, and treatment of these comorbid conditions is indicated.^36,43,52 For example, in relation to obesity, the use of metformin is very well tolerated and seems to positively impact HS symptoms.⁴³ Several studies have suggested that weight reduction lowers disease severity.^28-30 Patients should be counseled on the importance of smoking cessation and weight loss.

Finally, the emotional impact of HS is not to be discounted, both the physical and social discomfort as well as the chronicity of the disease and frustration with treatment.⁵¹ Chronic pain has been associated with increased rates of depression, and 43% of patients with HS specifically have been diagnosed with major depressive disorder.⁷ For these reasons, clinician guidance, social support, and websites can improve patient understanding of the disease, adherence to treatment, and comorbid anxiety and depression.⁵²

Topical Therapy

Topical therapy generally is limited to mild disease and is geared at decreasing inflammation or superimposed infection.^36,52 Some of the earliest therapies used were topical antibiotics.⁴³ Topical clindamycin has been shown to be as effective as oral tetracyclines in reducing the number of abscesses, but neither treatment substantially reduces pain associated with smaller nodules.⁵⁴ Intralesional corticosteroids such as triamcinolone acetonide have been shown to decrease both patient-reported pain and physician-assessed severity within 1 to 7 days.⁴² Routine injection, however, is not a feasible means of long-term treatment both because of inconvenience and the potential adverse effects of corticosteroids.^36,52 Both topical clindamycin and intralesional steroids are helpful in reducing inflammation prior to planned surgical intervention.^36,52,53

Newer topical therapies include resorcinol peels and combination antimicrobials, such as 2% triclosan and oral zinc gluconate.^52,53 Data surrounding the use of resorcinol in mild to moderate HS are promising and have shown decreased severity of both new and long-standing nodules. Fifteen-percent resorcinol peels are helpful tools that allow for self-administration by patients during exacerbations to decrease pain and flare duration.^55,56 In a 2016 clinical trial, a combination of oral zinc gluconate with topical triclosan was shown to reduce flare-ups and nodules in mild HS.⁵⁷ Oral zinc alone may have anti-inflammatory properties and generally is well tolerated.^43,53 Topical therapies have a role in reducing HS-associated pain but often are limited to milder disease.

Systemic Agents

Several therapeutic options exist for the treatment of HS; however, a detailed description of their mechanisms and efficacies is beyond the scope of this review, which is focused on pain. Briefly, these systemic agents include antibiotics, retinoids, corticosteroids, antiandrogens, and biologics.^43,52,53

Treatment with antibiotics such as tetracyclines or a combination of clindamycin plus rifampin has been shown to produce complete remission in 60% to 80% of users; however, this treatment requires more than 6 months of antibiotic therapy, which can be difficult to tolerate.^52,53,58 Relapse is common after antibiotic cessation.^2,43,52 Antibiotics have demonstrated efficacy during acute flares and in reducing inflammatory activity prior to surgery.⁵²

Retinoids have been utilized in the treatment of HS because of their action on sebaceous glands and hair follicles.^43,53 Acitretin has been shown to be the most effective oral retinoid available in the United States.⁴³ Unfortunately, many of the studies investigating the use of retinoids for treatment of HS are limited by small sample size.^36,43,52

Because HS is predominantly an inflammatory condition, immunosuppressants have been adapted to manage patients when antibiotics and topicals have failed. Systemic steroids rarely are used for long-term therapy because of the severe side effects and are preferred only for acute management.^36,52 Cyclosporine and dapsone have demonstrated efficacy in treating moderate to severe HS, whereas methotrexate and colchicine have shown little efficacy.⁵² Both cyclosporine and dapsone are difficult to tolerate, require laboratory monitoring, and lead to only conservative improvement rather than remission in most patients.⁴³

Immune dysregulation in HS involves elevated levels of proinflammatory cytokines such as tumor necrosis factor α (TNF-α), which is a key mediator of inflammation and a stimulator of other inflammatory cytokines.^59,60 The first approved biologic treatment of HS was adalimumab, a TNF-α inhibitor, which showed a 50% reduction in total abscess and inflammatory nodule count in 60% of patients with moderate to severe HS.^61-63 Of course, TNF-α inhibitor therapy is not without risks, specifically those of infection.^43,53,61,62 Maintenance therapy may be required if patients relapse.^53,61

Various interleukin inhibitors also have emerged as potential therapies for HS, such as ustekinumab and anakinra.^36,64 Both have been subject to numerous small case trials that have reported improvements in clinical severity and pain; however, both drugs were associated with a fair number of nonresponders.^36,64,65

Surgical Procedures

Although HS lesions may regress on their own in a matter of weeks, surgical drainage allows an acute alleviation of the severe burning pain associated with HS flares.^36,52,53 Because of improved understanding of the disease pathophysiology, recent therapies targeting the hair follicle have been developed and have shown promising results. These therapies include laser- and light-based procedures. Long-pulsed Nd:YAG laser therapy reduces the number of hair follicles and sebaceous glands and has been effective for Hurley stage I or II disease.^{36,43,52,53,66} Photodynamic therapy offers a less-invasive option compared to surgery and laser therapy.^52,53,66 Both Nd:YAG and CO₂ laser therapy offer low recurrence rates (<30%) due to destruction of the apocrine unit.^43,53 Photodynamic therapy for mild disease offers a less-invasive option compared to surgery and laser therapy.⁵³ There is a need for larger randomized controlled trials involving laser, light, and CO₂ therapies.⁶⁶

Conclusion

Hidradenitis suppurativa is a debilitating condition with an underestimated disease burden. Although the pathophysiology of the disease is not completely understood, it is evident that pain is a major cause of morbidity. Patients experience a multitude of acute and chronic pain types: inflammatory, noninflammatory, nociceptive, neuropathic, and ischemic. Pain perception and quality of life are further impacted by psychiatric conditions such as depression and anxiety, both of which are common comorbidities in patients with HS. Several pharmacologic agents have been used to treat HS-associated pain with mixed results. First-line treatment of acute pain episodes includes oral acetaminophen, NSAIDs, and topical analgesics. Management of chronic pain includes utilization of topical agents, systemic agents, and biologics, as well as addressing lifestyle (eg, obesity, smoking status) and psychiatric comorbidities. Although these therapies have roles in HS pain management, the most effective pain remedies developed thus far are limited to surgery and TNF-α inhibitors. Optimization of pain control in patients with HS requires multidisciplinary collaboration among dermatologists, pain specialists, psychiatrists, and other members of the health care team. Further large-scale studies are needed to create an evidence-based treatment algorithm for the management of pain in HS.

Hidradenitis suppurativa (HS) is a chronic inflammatory, androgen gland disorder characterized by recurrent rupture of the hair follicles with a vigorous inflammatory response. This response results in abscess formation and development of draining sinus tracts and hypertrophic fibrous scars.^1,2 Pain, discomfort, and odorous discharge from the recalcitrant lesions have a profound impact on patient quality of life.^3,4

The morbidity and disease burden associated with HS are particularly underestimated, as patients frequently report debilitating pain that often is overlooked.^5,6 Additionally, the quality and intensity of perceived pain are compounded by frequently associated depression and anxiety.^7-9 Pain has been reported by patients with HS to be the highest cause of morbidity, despite the disfiguring nature of the disease and its associated psychosocial distress.^7,10 Nonetheless, HS lacks an accepted pain management algorithm similar to those that have been developed for the treatment of other acute or chronic pain disorders, such as back pain and sickle cell disease.^4,11-13

Given the lack of formal studies regarding pain management in patients with HS, clinicians are limited to general pain guidelines, expert opinion, small trials, and patient preference.³ Furthermore, effective pain management in HS necessitates the treatment of both chronic pain affecting daily function and acute pain present during disease flares, surgical interventions, and dressing changes.³ The result is a wide array of strategies used for HS-associated pain.^3,4

Epidemiology and Pathophysiology

Hidradenitis suppurativa historically has been an overlooked and underdiagnosed disease, which limits epidemiology data.⁵ Current estimates are that HS affects approximately 1% of the general population; however, prevalence rates range from 0.03% to 4.1%.^14-16

The exact etiology of HS remains unclear, but it is thought that genetic factors, immune dysregulation, and environmental/behavioral influences all contribute to its pathophysiology.^1,17 Up to 40% of patients with HS report a positive family history of the disease.^18-20 Hidradenitis suppurativa has been associated with other inflammatory disease states, such as inflammatory bowel disease, spondyloarthropathies, and pyoderma gangrenosum.^16,21,22

It is thought that HS is the result of some defect in keratin clearance that leads to follicular hyperkeratinization and occlusion.¹ Resultant rupture of pilosebaceous units and spillage of contents (including keratin and bacteria) into the surrounding dermis triggers a vigorous inflammatory response. Sinus tracts and fistulas become the targets of bacterial colonization, biofilm formation, and secondary infection. The result is suppuration and extension of the lesions as well as sustained chronic inflammation.^23,24

Although the etiology of HS is complex, several modifiable risk factors for the disease have been identified, most prominently cigarette smoking and obesity. Approximately 70% of patients with HS smoke cigarettes.^2,15,25,26 Obesity has a well-known association with HS, and it is possible that weight reduction lowers disease severity.^27-30

Clinical Presentation and Diagnosis

Establishing a diagnosis of HS necessitates recognition of disease morphology, topography, and chronicity. Hidradenitis suppurativa most commonly occurs in the axillae, inguinal and anogenital region, perineal region, and inframammary region.^5,31 A typical history involves a prolonged disease course with recurrent lesions and intermittent periods of improvement or remission. Primary lesions are deep, inflamed, painful, and sterile. Ultimately, these lesions rupture and track subcutaneously.^15,25 Intercommunicating sinus tracts form from multiple recurrent nodules in close proximity and may ultimately lead to fibrotic scarring and local architectural distortion.³² The Hurley staging system helps to guide treatment interventions based on disease severity. Approach to pain management is discussed below.

Pain Management in HS: General Principles

Pain management is complex for clinicians, as there are limited studies from which to draw treatment recommendations. Incomplete understanding of the etiology and pathophysiology of the disease contributes to the lack of established management guidelines.

A PubMed search of articles indexed for MEDLINE using the terms hidradenitis, suppurativa, pain, and management revealed 61 different results dating back to 1980, 52 of which had been published in the last 5 years. When the word acute was added to the search, there were only 6 results identified. These results clearly reflect a better understanding of HS-mediated pain as well as clinical unmet needs and evolving strategies in pain management therapeutics. However, many of these studies reflect therapies focused on the mediation or modulation of HS pathogenesis rather than potential pain management therapies.

In addition, the heterogenous nature of the pain experience in HS poses a challenge for clinicians. Patients may experience multiple pain types concurrently, including inflammatory, noninflammatory, nociceptive, neuropathic, and ischemic, as well as pain related to arthritis.^3,33,34 Pain perception is further complicated by the observation that patients with HS have high rates of psychiatric comorbidities such as depression and anxiety, both of which profoundly alter perception of both the strength and quality of pain.^7,8,22,35 A suggested algorithm for treatment of pain in HS is described in the eTable.³⁶

Chronicity is a hallmark of HS. Patients experience a prolonged disease course involving acute painful exacerbations superimposed on chronic pain that affects all aspects of daily life. Changes in self-perception, daily living activities, mood state, physical functioning, and physical comfort frequently are reported to have a major impact on quality of life.^1,3,37

In 2018, Thorlacius et al³⁸ created a multistakeholder consensus on a core outcome set of domains detailing what to measure in clinical trials for HS. The authors hoped that the routine adoption of these core domains would promote the collection of consistent and relevant information, bolster the strength of evidence synthesis, and minimize the risk for outcome reporting bias among studies.³⁸ It is important to ascertain the patient’s description of his/her pain to distinguish between stimulus-dependent nociceptive pain vs spontaneous neuropathic pain.^3,7,10 The most common pain descriptors used by patients are “shooting,” “itchy,” “blinding,” “cutting,” and “exhausting.”¹⁰ In addition to obtaining descriptive factors, it is important for the clinician to obtain information on the timing of the pain, whether or not the pain is relieved with spontaneous or surgical drainage, and if the patient is experiencing chronic background pain secondary to scarring or skin contraction.³ With the routine utilization of a consistent set of core domains, advances in our understanding of the different elements of HS pain, and increased provider awareness of the disease, the future of pain management in patients with HS seems promising.

Acute and Perioperative Pain Management

Acute Pain Management—The pain in HS can range from mild to excruciating.^3,7 The difference between acute and chronic pain in this condition may be hard to delineate, as patients may have intense acute flares on top of a baseline level of chronic pain.^3,7,14 These factors, in combination with various pain types of differing etiologies, make the treatment of HS-associated pain a therapeutic challenge.

The first-line treatments for acute pain in HS are oral acetaminophen, oral nonsteroidal anti-inflammatory drugs (NSAIDs), and topical analgesics.³ These treatment modalities are especially helpful for nociceptive pain, which often is described as having an aching or tender quality.³ Topical treatment for acute pain episodes includes diclofenac gel and liposomal lidocaine cream.³⁹ Topical lidocaine in particular has the benefit of being rapid acting, and its effect can last 1 to 2 hours. Ketamine has been anecdotally used as a topical treatment. Treatment options for neuropathic pain include topical amitriptyline, gabapentin, and pregabalin.³⁹ Dressings and ice packs may be used in cases of mild acute pain, depending on patient preference.³

First-line therapies may not provide adequate pain control in many patients.^3,40,41 Should the first-line treatments fail, oral opiates can be considered as a treatment option, especially if the patient has a history of recurrent pain unresponsive to milder methods of pain control.^3,40,41 However, prudence should be exercised, as patients with HS have a higher risk for opioid abuse, and referral to a pain specialist is advisable.⁴⁰ Generally, use of opioids should be limited to the smallest period of time possible.^40,41 Codeine can be used as a first opioid option, with hydromorphone available as an alternative.⁴¹

Pain caused by inflamed abscesses and nodules can be treated with either intralesional corticosteroids or incision and drainage. Intralesional triamcinolone has been found to cause substantial pain relief within 1 day of injection in patients with HS.^3,42

Prompt discussion about the remitting course of HS will prepare patients for flares. Although the therapies discussed here aim to reduce the clinical severity and inflammation associated with HS, achieving pain-free remission can be challenging. Barriers to developing a long-term treatment regimen include intolerable side effects or simply nonresponsive disease.^36,43

Management of Perioperative Pain—Medical treatment of HS often yields only transient or mild results. Hurley stage II or III lesions typically require surgical removal of affected tissues.^32,44-46 Surgery may dramatically reduce the primary disease burden and provide substantial pain relief.^3,4,44 Complete resection of the affected tissue by wide excision is the most common surgical procedure used.^46-48 However, various tissue-sparing techniques, such as skin-tissue-sparing excision with electrosurgical peeling, also have been utilized. Tissue-sparing surgical techniques may lead to shorter healing times and less postoperative pain.⁴⁸

There currently is little guidance available on the perioperative management of pain as it relates to surgical procedures for HS. The pain experienced from surgery varies based on the area and location of affected tissue; extent of disease; surgical technique used; and whether primary closure, closure by secondary intention, or skin grafting is utilized.^47,49 Medical treatment aimed at reducing inflammation prior to surgical intervention may improve postoperative pain and complications.

The use of general vs local anesthesia during surgery depends on the extent of the disease and the amount of tissue being removed; however, the use of local anesthesia has been associated with a higher recurrence of disease, possibly owing to less aggressive tissue removal.⁵⁰ Intraoperatively, the injection of 0.5% bupivacaine around the wound edges may lead to less postoperative pain.^3,48 Postoperative pain usually is managed with acetaminophen and NSAIDs.⁴⁸ In cases of severe postoperative pain, short- and long-acting opioid oxycodone preparations may be used. The combination of diclofenac and tramadol also has been used postoperatively.³ Patients who do not undergo extensive surgery often can leave the hospital the same day.

Effective strategies for mitigating HS-associated pain must address the chronic pain component of the disease. Long-term management involves lifestyle modifications and pharmacologic agents.

Chronic Pain Management

Although HS is not a curable disease, there are treatments available to minimize symptoms. Long-term management of HS is essential to minimize the effects of chronic pain and physical scarring associated with inflammation.³¹ In one study from the French Society of Dermatology, pain reported by patients with HS was directly associated with severity and duration of disease, emotional symptoms, and reduced functionality.⁵¹ For these reasons, many treatments for HS target reducing clinical severity and achieving remission, often defined as more than 6 months without any recurrence of lesions.⁵² In addition to lifestyle management, therapies available to manage HS include topical and systemic medications as well as procedures such as surgical excision.^36,43,52,53

Lifestyle Modifications

Regardless of the severity of HS, all patients may benefit from basic education on the pathogenesis of the disease.³⁶ The associations with smoking and obesity have been well documented, and treatment of these comorbid conditions is indicated.^36,43,52 For example, in relation to obesity, the use of metformin is very well tolerated and seems to positively impact HS symptoms.⁴³ Several studies have suggested that weight reduction lowers disease severity.^28-30 Patients should be counseled on the importance of smoking cessation and weight loss.

Finally, the emotional impact of HS is not to be discounted, both the physical and social discomfort as well as the chronicity of the disease and frustration with treatment.⁵¹ Chronic pain has been associated with increased rates of depression, and 43% of patients with HS specifically have been diagnosed with major depressive disorder.⁷ For these reasons, clinician guidance, social support, and websites can improve patient understanding of the disease, adherence to treatment, and comorbid anxiety and depression.⁵²

Topical Therapy

Topical therapy generally is limited to mild disease and is geared at decreasing inflammation or superimposed infection.^36,52 Some of the earliest therapies used were topical antibiotics.⁴³ Topical clindamycin has been shown to be as effective as oral tetracyclines in reducing the number of abscesses, but neither treatment substantially reduces pain associated with smaller nodules.⁵⁴ Intralesional corticosteroids such as triamcinolone acetonide have been shown to decrease both patient-reported pain and physician-assessed severity within 1 to 7 days.⁴² Routine injection, however, is not a feasible means of long-term treatment both because of inconvenience and the potential adverse effects of corticosteroids.^36,52 Both topical clindamycin and intralesional steroids are helpful in reducing inflammation prior to planned surgical intervention.^36,52,53

Newer topical therapies include resorcinol peels and combination antimicrobials, such as 2% triclosan and oral zinc gluconate.^52,53 Data surrounding the use of resorcinol in mild to moderate HS are promising and have shown decreased severity of both new and long-standing nodules. Fifteen-percent resorcinol peels are helpful tools that allow for self-administration by patients during exacerbations to decrease pain and flare duration.^55,56 In a 2016 clinical trial, a combination of oral zinc gluconate with topical triclosan was shown to reduce flare-ups and nodules in mild HS.⁵⁷ Oral zinc alone may have anti-inflammatory properties and generally is well tolerated.^43,53 Topical therapies have a role in reducing HS-associated pain but often are limited to milder disease.

Systemic Agents

Several therapeutic options exist for the treatment of HS; however, a detailed description of their mechanisms and efficacies is beyond the scope of this review, which is focused on pain. Briefly, these systemic agents include antibiotics, retinoids, corticosteroids, antiandrogens, and biologics.^43,52,53

Treatment with antibiotics such as tetracyclines or a combination of clindamycin plus rifampin has been shown to produce complete remission in 60% to 80% of users; however, this treatment requires more than 6 months of antibiotic therapy, which can be difficult to tolerate.^52,53,58 Relapse is common after antibiotic cessation.^2,43,52 Antibiotics have demonstrated efficacy during acute flares and in reducing inflammatory activity prior to surgery.⁵²

Retinoids have been utilized in the treatment of HS because of their action on sebaceous glands and hair follicles.^43,53 Acitretin has been shown to be the most effective oral retinoid available in the United States.⁴³ Unfortunately, many of the studies investigating the use of retinoids for treatment of HS are limited by small sample size.^36,43,52

Because HS is predominantly an inflammatory condition, immunosuppressants have been adapted to manage patients when antibiotics and topicals have failed. Systemic steroids rarely are used for long-term therapy because of the severe side effects and are preferred only for acute management.^36,52 Cyclosporine and dapsone have demonstrated efficacy in treating moderate to severe HS, whereas methotrexate and colchicine have shown little efficacy.⁵² Both cyclosporine and dapsone are difficult to tolerate, require laboratory monitoring, and lead to only conservative improvement rather than remission in most patients.⁴³

Immune dysregulation in HS involves elevated levels of proinflammatory cytokines such as tumor necrosis factor α (TNF-α), which is a key mediator of inflammation and a stimulator of other inflammatory cytokines.^59,60 The first approved biologic treatment of HS was adalimumab, a TNF-α inhibitor, which showed a 50% reduction in total abscess and inflammatory nodule count in 60% of patients with moderate to severe HS.^61-63 Of course, TNF-α inhibitor therapy is not without risks, specifically those of infection.^43,53,61,62 Maintenance therapy may be required if patients relapse.^53,61

Various interleukin inhibitors also have emerged as potential therapies for HS, such as ustekinumab and anakinra.^36,64 Both have been subject to numerous small case trials that have reported improvements in clinical severity and pain; however, both drugs were associated with a fair number of nonresponders.^36,64,65

Surgical Procedures

Although HS lesions may regress on their own in a matter of weeks, surgical drainage allows an acute alleviation of the severe burning pain associated with HS flares.^36,52,53 Because of improved understanding of the disease pathophysiology, recent therapies targeting the hair follicle have been developed and have shown promising results. These therapies include laser- and light-based procedures. Long-pulsed Nd:YAG laser therapy reduces the number of hair follicles and sebaceous glands and has been effective for Hurley stage I or II disease.^{36,43,52,53,66} Photodynamic therapy offers a less-invasive option compared to surgery and laser therapy.^52,53,66 Both Nd:YAG and CO₂ laser therapy offer low recurrence rates (<30%) due to destruction of the apocrine unit.^43,53 Photodynamic therapy for mild disease offers a less-invasive option compared to surgery and laser therapy.⁵³ There is a need for larger randomized controlled trials involving laser, light, and CO₂ therapies.⁶⁶

Conclusion

Hidradenitis suppurativa is a debilitating condition with an underestimated disease burden. Although the pathophysiology of the disease is not completely understood, it is evident that pain is a major cause of morbidity. Patients experience a multitude of acute and chronic pain types: inflammatory, noninflammatory, nociceptive, neuropathic, and ischemic. Pain perception and quality of life are further impacted by psychiatric conditions such as depression and anxiety, both of which are common comorbidities in patients with HS. Several pharmacologic agents have been used to treat HS-associated pain with mixed results. First-line treatment of acute pain episodes includes oral acetaminophen, NSAIDs, and topical analgesics. Management of chronic pain includes utilization of topical agents, systemic agents, and biologics, as well as addressing lifestyle (eg, obesity, smoking status) and psychiatric comorbidities. Although these therapies have roles in HS pain management, the most effective pain remedies developed thus far are limited to surgery and TNF-α inhibitors. Optimization of pain control in patients with HS requires multidisciplinary collaboration among dermatologists, pain specialists, psychiatrists, and other members of the health care team. Further large-scale studies are needed to create an evidence-based treatment algorithm for the management of pain in HS.

Hidradenitis suppurativa (HS) is a chronic inflammatory, androgen gland disorder characterized by recurrent rupture of the hair follicles with a vigorous inflammatory response. This response results in abscess formation and development of draining sinus tracts and hypertrophic fibrous scars.^1,2 Pain, discomfort, and odorous discharge from the recalcitrant lesions have a profound impact on patient quality of life.^3,4

The morbidity and disease burden associated with HS are particularly underestimated, as patients frequently report debilitating pain that often is overlooked.^5,6 Additionally, the quality and intensity of perceived pain are compounded by frequently associated depression and anxiety.^7-9 Pain has been reported by patients with HS to be the highest cause of morbidity, despite the disfiguring nature of the disease and its associated psychosocial distress.^7,10 Nonetheless, HS lacks an accepted pain management algorithm similar to those that have been developed for the treatment of other acute or chronic pain disorders, such as back pain and sickle cell disease.^4,11-13

Given the lack of formal studies regarding pain management in patients with HS, clinicians are limited to general pain guidelines, expert opinion, small trials, and patient preference.³ Furthermore, effective pain management in HS necessitates the treatment of both chronic pain affecting daily function and acute pain present during disease flares, surgical interventions, and dressing changes.³ The result is a wide array of strategies used for HS-associated pain.^3,4

Epidemiology and Pathophysiology

Hidradenitis suppurativa historically has been an overlooked and underdiagnosed disease, which limits epidemiology data.⁵ Current estimates are that HS affects approximately 1% of the general population; however, prevalence rates range from 0.03% to 4.1%.^14-16

The exact etiology of HS remains unclear, but it is thought that genetic factors, immune dysregulation, and environmental/behavioral influences all contribute to its pathophysiology.^1,17 Up to 40% of patients with HS report a positive family history of the disease.^18-20 Hidradenitis suppurativa has been associated with other inflammatory disease states, such as inflammatory bowel disease, spondyloarthropathies, and pyoderma gangrenosum.^16,21,22

It is thought that HS is the result of some defect in keratin clearance that leads to follicular hyperkeratinization and occlusion.¹ Resultant rupture of pilosebaceous units and spillage of contents (including keratin and bacteria) into the surrounding dermis triggers a vigorous inflammatory response. Sinus tracts and fistulas become the targets of bacterial colonization, biofilm formation, and secondary infection. The result is suppuration and extension of the lesions as well as sustained chronic inflammation.^23,24

Although the etiology of HS is complex, several modifiable risk factors for the disease have been identified, most prominently cigarette smoking and obesity. Approximately 70% of patients with HS smoke cigarettes.^2,15,25,26 Obesity has a well-known association with HS, and it is possible that weight reduction lowers disease severity.^27-30

Clinical Presentation and Diagnosis

Establishing a diagnosis of HS necessitates recognition of disease morphology, topography, and chronicity. Hidradenitis suppurativa most commonly occurs in the axillae, inguinal and anogenital region, perineal region, and inframammary region.^5,31 A typical history involves a prolonged disease course with recurrent lesions and intermittent periods of improvement or remission. Primary lesions are deep, inflamed, painful, and sterile. Ultimately, these lesions rupture and track subcutaneously.^15,25 Intercommunicating sinus tracts form from multiple recurrent nodules in close proximity and may ultimately lead to fibrotic scarring and local architectural distortion.³² The Hurley staging system helps to guide treatment interventions based on disease severity. Approach to pain management is discussed below.

Pain Management in HS: General Principles

Pain management is complex for clinicians, as there are limited studies from which to draw treatment recommendations. Incomplete understanding of the etiology and pathophysiology of the disease contributes to the lack of established management guidelines.

A PubMed search of articles indexed for MEDLINE using the terms hidradenitis, suppurativa, pain, and management revealed 61 different results dating back to 1980, 52 of which had been published in the last 5 years. When the word acute was added to the search, there were only 6 results identified. These results clearly reflect a better understanding of HS-mediated pain as well as clinical unmet needs and evolving strategies in pain management therapeutics. However, many of these studies reflect therapies focused on the mediation or modulation of HS pathogenesis rather than potential pain management therapies.

In addition, the heterogenous nature of the pain experience in HS poses a challenge for clinicians. Patients may experience multiple pain types concurrently, including inflammatory, noninflammatory, nociceptive, neuropathic, and ischemic, as well as pain related to arthritis.^3,33,34 Pain perception is further complicated by the observation that patients with HS have high rates of psychiatric comorbidities such as depression and anxiety, both of which profoundly alter perception of both the strength and quality of pain.^7,8,22,35 A suggested algorithm for treatment of pain in HS is described in the eTable.³⁶

Chronicity is a hallmark of HS. Patients experience a prolonged disease course involving acute painful exacerbations superimposed on chronic pain that affects all aspects of daily life. Changes in self-perception, daily living activities, mood state, physical functioning, and physical comfort frequently are reported to have a major impact on quality of life.^1,3,37

In 2018, Thorlacius et al³⁸ created a multistakeholder consensus on a core outcome set of domains detailing what to measure in clinical trials for HS. The authors hoped that the routine adoption of these core domains would promote the collection of consistent and relevant information, bolster the strength of evidence synthesis, and minimize the risk for outcome reporting bias among studies.³⁸ It is important to ascertain the patient’s description of his/her pain to distinguish between stimulus-dependent nociceptive pain vs spontaneous neuropathic pain.^3,7,10 The most common pain descriptors used by patients are “shooting,” “itchy,” “blinding,” “cutting,” and “exhausting.”¹⁰ In addition to obtaining descriptive factors, it is important for the clinician to obtain information on the timing of the pain, whether or not the pain is relieved with spontaneous or surgical drainage, and if the patient is experiencing chronic background pain secondary to scarring or skin contraction.³ With the routine utilization of a consistent set of core domains, advances in our understanding of the different elements of HS pain, and increased provider awareness of the disease, the future of pain management in patients with HS seems promising.

Acute and Perioperative Pain Management

Acute Pain Management—The pain in HS can range from mild to excruciating.^3,7 The difference between acute and chronic pain in this condition may be hard to delineate, as patients may have intense acute flares on top of a baseline level of chronic pain.^3,7,14 These factors, in combination with various pain types of differing etiologies, make the treatment of HS-associated pain a therapeutic challenge.

The first-line treatments for acute pain in HS are oral acetaminophen, oral nonsteroidal anti-inflammatory drugs (NSAIDs), and topical analgesics.³ These treatment modalities are especially helpful for nociceptive pain, which often is described as having an aching or tender quality.³ Topical treatment for acute pain episodes includes diclofenac gel and liposomal lidocaine cream.³⁹ Topical lidocaine in particular has the benefit of being rapid acting, and its effect can last 1 to 2 hours. Ketamine has been anecdotally used as a topical treatment. Treatment options for neuropathic pain include topical amitriptyline, gabapentin, and pregabalin.³⁹ Dressings and ice packs may be used in cases of mild acute pain, depending on patient preference.³

First-line therapies may not provide adequate pain control in many patients.^3,40,41 Should the first-line treatments fail, oral opiates can be considered as a treatment option, especially if the patient has a history of recurrent pain unresponsive to milder methods of pain control.^3,40,41 However, prudence should be exercised, as patients with HS have a higher risk for opioid abuse, and referral to a pain specialist is advisable.⁴⁰ Generally, use of opioids should be limited to the smallest period of time possible.^40,41 Codeine can be used as a first opioid option, with hydromorphone available as an alternative.⁴¹

Pain caused by inflamed abscesses and nodules can be treated with either intralesional corticosteroids or incision and drainage. Intralesional triamcinolone has been found to cause substantial pain relief within 1 day of injection in patients with HS.^3,42

Prompt discussion about the remitting course of HS will prepare patients for flares. Although the therapies discussed here aim to reduce the clinical severity and inflammation associated with HS, achieving pain-free remission can be challenging. Barriers to developing a long-term treatment regimen include intolerable side effects or simply nonresponsive disease.^36,43

Management of Perioperative Pain—Medical treatment of HS often yields only transient or mild results. Hurley stage II or III lesions typically require surgical removal of affected tissues.^32,44-46 Surgery may dramatically reduce the primary disease burden and provide substantial pain relief.^3,4,44 Complete resection of the affected tissue by wide excision is the most common surgical procedure used.^46-48 However, various tissue-sparing techniques, such as skin-tissue-sparing excision with electrosurgical peeling, also have been utilized. Tissue-sparing surgical techniques may lead to shorter healing times and less postoperative pain.⁴⁸

There currently is little guidance available on the perioperative management of pain as it relates to surgical procedures for HS. The pain experienced from surgery varies based on the area and location of affected tissue; extent of disease; surgical technique used; and whether primary closure, closure by secondary intention, or skin grafting is utilized.^47,49 Medical treatment aimed at reducing inflammation prior to surgical intervention may improve postoperative pain and complications.

The use of general vs local anesthesia during surgery depends on the extent of the disease and the amount of tissue being removed; however, the use of local anesthesia has been associated with a higher recurrence of disease, possibly owing to less aggressive tissue removal.⁵⁰ Intraoperatively, the injection of 0.5% bupivacaine around the wound edges may lead to less postoperative pain.^3,48 Postoperative pain usually is managed with acetaminophen and NSAIDs.⁴⁸ In cases of severe postoperative pain, short- and long-acting opioid oxycodone preparations may be used. The combination of diclofenac and tramadol also has been used postoperatively.³ Patients who do not undergo extensive surgery often can leave the hospital the same day.

Effective strategies for mitigating HS-associated pain must address the chronic pain component of the disease. Long-term management involves lifestyle modifications and pharmacologic agents.

Chronic Pain Management

Although HS is not a curable disease, there are treatments available to minimize symptoms. Long-term management of HS is essential to minimize the effects of chronic pain and physical scarring associated with inflammation.³¹ In one study from the French Society of Dermatology, pain reported by patients with HS was directly associated with severity and duration of disease, emotional symptoms, and reduced functionality.⁵¹ For these reasons, many treatments for HS target reducing clinical severity and achieving remission, often defined as more than 6 months without any recurrence of lesions.⁵² In addition to lifestyle management, therapies available to manage HS include topical and systemic medications as well as procedures such as surgical excision.^36,43,52,53

Lifestyle Modifications

Regardless of the severity of HS, all patients may benefit from basic education on the pathogenesis of the disease.³⁶ The associations with smoking and obesity have been well documented, and treatment of these comorbid conditions is indicated.^36,43,52 For example, in relation to obesity, the use of metformin is very well tolerated and seems to positively impact HS symptoms.⁴³ Several studies have suggested that weight reduction lowers disease severity.^28-30 Patients should be counseled on the importance of smoking cessation and weight loss.

Finally, the emotional impact of HS is not to be discounted, both the physical and social discomfort as well as the chronicity of the disease and frustration with treatment.⁵¹ Chronic pain has been associated with increased rates of depression, and 43% of patients with HS specifically have been diagnosed with major depressive disorder.⁷ For these reasons, clinician guidance, social support, and websites can improve patient understanding of the disease, adherence to treatment, and comorbid anxiety and depression.⁵²

Topical Therapy

Topical therapy generally is limited to mild disease and is geared at decreasing inflammation or superimposed infection.^36,52 Some of the earliest therapies used were topical antibiotics.⁴³ Topical clindamycin has been shown to be as effective as oral tetracyclines in reducing the number of abscesses, but neither treatment substantially reduces pain associated with smaller nodules.⁵⁴ Intralesional corticosteroids such as triamcinolone acetonide have been shown to decrease both patient-reported pain and physician-assessed severity within 1 to 7 days.⁴² Routine injection, however, is not a feasible means of long-term treatment both because of inconvenience and the potential adverse effects of corticosteroids.^36,52 Both topical clindamycin and intralesional steroids are helpful in reducing inflammation prior to planned surgical intervention.^36,52,53

Newer topical therapies include resorcinol peels and combination antimicrobials, such as 2% triclosan and oral zinc gluconate.^52,53 Data surrounding the use of resorcinol in mild to moderate HS are promising and have shown decreased severity of both new and long-standing nodules. Fifteen-percent resorcinol peels are helpful tools that allow for self-administration by patients during exacerbations to decrease pain and flare duration.^55,56 In a 2016 clinical trial, a combination of oral zinc gluconate with topical triclosan was shown to reduce flare-ups and nodules in mild HS.⁵⁷ Oral zinc alone may have anti-inflammatory properties and generally is well tolerated.^43,53 Topical therapies have a role in reducing HS-associated pain but often are limited to milder disease.

Systemic Agents

Several therapeutic options exist for the treatment of HS; however, a detailed description of their mechanisms and efficacies is beyond the scope of this review, which is focused on pain. Briefly, these systemic agents include antibiotics, retinoids, corticosteroids, antiandrogens, and biologics.^43,52,53

Treatment with antibiotics such as tetracyclines or a combination of clindamycin plus rifampin has been shown to produce complete remission in 60% to 80% of users; however, this treatment requires more than 6 months of antibiotic therapy, which can be difficult to tolerate.^52,53,58 Relapse is common after antibiotic cessation.^2,43,52 Antibiotics have demonstrated efficacy during acute flares and in reducing inflammatory activity prior to surgery.⁵²

Retinoids have been utilized in the treatment of HS because of their action on sebaceous glands and hair follicles.^43,53 Acitretin has been shown to be the most effective oral retinoid available in the United States.⁴³ Unfortunately, many of the studies investigating the use of retinoids for treatment of HS are limited by small sample size.^36,43,52

Because HS is predominantly an inflammatory condition, immunosuppressants have been adapted to manage patients when antibiotics and topicals have failed. Systemic steroids rarely are used for long-term therapy because of the severe side effects and are preferred only for acute management.^36,52 Cyclosporine and dapsone have demonstrated efficacy in treating moderate to severe HS, whereas methotrexate and colchicine have shown little efficacy.⁵² Both cyclosporine and dapsone are difficult to tolerate, require laboratory monitoring, and lead to only conservative improvement rather than remission in most patients.⁴³

Immune dysregulation in HS involves elevated levels of proinflammatory cytokines such as tumor necrosis factor α (TNF-α), which is a key mediator of inflammation and a stimulator of other inflammatory cytokines.^59,60 The first approved biologic treatment of HS was adalimumab, a TNF-α inhibitor, which showed a 50% reduction in total abscess and inflammatory nodule count in 60% of patients with moderate to severe HS.^61-63 Of course, TNF-α inhibitor therapy is not without risks, specifically those of infection.^43,53,61,62 Maintenance therapy may be required if patients relapse.^53,61

Various interleukin inhibitors also have emerged as potential therapies for HS, such as ustekinumab and anakinra.^36,64 Both have been subject to numerous small case trials that have reported improvements in clinical severity and pain; however, both drugs were associated with a fair number of nonresponders.^36,64,65

Surgical Procedures

Although HS lesions may regress on their own in a matter of weeks, surgical drainage allows an acute alleviation of the severe burning pain associated with HS flares.^36,52,53 Because of improved understanding of the disease pathophysiology, recent therapies targeting the hair follicle have been developed and have shown promising results. These therapies include laser- and light-based procedures. Long-pulsed Nd:YAG laser therapy reduces the number of hair follicles and sebaceous glands and has been effective for Hurley stage I or II disease.^{36,43,52,53,66} Photodynamic therapy offers a less-invasive option compared to surgery and laser therapy.^52,53,66 Both Nd:YAG and CO₂ laser therapy offer low recurrence rates (<30%) due to destruction of the apocrine unit.^43,53 Photodynamic therapy for mild disease offers a less-invasive option compared to surgery and laser therapy.⁵³ There is a need for larger randomized controlled trials involving laser, light, and CO₂ therapies.⁶⁶

Conclusion

Hidradenitis suppurativa is a debilitating condition with an underestimated disease burden. Although the pathophysiology of the disease is not completely understood, it is evident that pain is a major cause of morbidity. Patients experience a multitude of acute and chronic pain types: inflammatory, noninflammatory, nociceptive, neuropathic, and ischemic. Pain perception and quality of life are further impacted by psychiatric conditions such as depression and anxiety, both of which are common comorbidities in patients with HS. Several pharmacologic agents have been used to treat HS-associated pain with mixed results. First-line treatment of acute pain episodes includes oral acetaminophen, NSAIDs, and topical analgesics. Management of chronic pain includes utilization of topical agents, systemic agents, and biologics, as well as addressing lifestyle (eg, obesity, smoking status) and psychiatric comorbidities. Although these therapies have roles in HS pain management, the most effective pain remedies developed thus far are limited to surgery and TNF-α inhibitors. Optimization of pain control in patients with HS requires multidisciplinary collaboration among dermatologists, pain specialists, psychiatrists, and other members of the health care team. Further large-scale studies are needed to create an evidence-based treatment algorithm for the management of pain in HS.

The collective struggle felt by healthcare workers simultaneously learning about and caring for patients impacted by SARS-CoV2 infections throughout 2020 was physically and emotionally exhausting. The majority of us had never experienced a global pandemic. Beyond our work in the professional arena of ambulatory practices and hospitals, we also felt the soul-crushing impact of the pandemic in every other aspect of our lives. Preexisting health disparities were amplified by COVID-19. Some of the most affected communities also bore the weight of an additional tsunami of ongoing racial injustice.¹ And as healthcare workers, we did our best to process and navigate it all while trying to avoid burnout—as well as being infected with COVID-19 ourselves. When the news of the highly effective vaccines against SARS-CoV2 receiving emergency use authorization broke late in 2020, it felt like a light at the end of a very dark tunnel.

In the weeks preceding wide availability of the vaccines, it became apparent that significant numbers of people lacked confidence in the vaccines. Given the disproportionate impact of COVID-19 on racial minorities, much of the discussion centered around “vaccine hesitancy” in these communities. Reasons such as historical mistrust, belief in conspiracy theories, and misinformation emerged as the leading explanations.² Campaigns and educational programs targeting Black Americans were quickly developed to counter this widely distributed narrative.

Vaccine uptake also became politicized, which created additional challenges. As schools and businesses reopened, the voices of those opposing pandemic mitigation mandates such as masking and vaccination were highlighted by media outlets. And though a large movement of individuals who had opted against vaccines existed well before the pandemic, with few exceptions, that number had never been great enough to impact public health to this extent.³ This primarily nonminority group of unvaccinated individuals also morphed into another monolithic identity: the “anti-vaxxer.”

The lion’s share of discussions around vaccine uptake centered on these two groups: the “vaccine hesitant” minority and the “anti-vaxxer.” The perspectives and frustration around these two stereotypical unvaccinated groups were underscored in journals and the lay press. But those working in communities and in direct care came into contact with countless COVID-19-positive patients who were unvaccinated and fell into neither of these categories. There was a large swath of vulnerable people who still had unanswered questions and mistrust in the medical system standing in their way. Awareness of health disparities among racial minorities is something that was discussed among providers, but it was something experienced and felt by patients daily in regard to so much more than just COVID-19.

With broader access to vaccines through retail, community-based, and clinical facilities, more patients who desired vaccination had the opportunity. After an initial rise in vaccine uptake, the numbers plateaued. But what remained was the repetitive messaging and sustained focus directed toward Black people and their “vaccine hesitancy.”

Grady Memorial Hospital, a public safety net hospital in Atlanta, serves a predominantly Black and uninsured patient population. We found that a “FAQ” approach with a narrow range of hypothetical ideas about unvaccinated minorities clashed with the reality of what we encountered in clinical environments and the community. While misinformation did appear to be prevalent, we appreciated that the context and level of conviction were heterogenous. We appreciated that each individual conversation could reveal something new to us about that unique patient and their personal concerns about vaccination. As time moved forward, it became clear that there was no playbook for any group, especially for historically disadvantaged communities. Importantly, it was recognized that attempts to anticipate what may be a person’s barrier to vaccination often worked to further erode trust. However, when we focused on creating a space for dialogue, we found we were able to move beyond information-sharing and instead were able to co-construct interpretations of information and co-create solutions that matched patients’ values and lived experiences.⁴ Through dialogue, we were better able to be transparent about our own experiences, which ultimately facilitated authentic conversations with patients.

In September 2021, we approached our hospital leadership with a patient-centered strategy aimed at providing our patients, staff, and visitors a psychologically safe place to discuss vaccine-related concerns without judgment. With their support, we set up a table in the busiest part of our hospital atrium between the information desk and vaccine-administration site. Beside it was a folding board sign with an image and these words:

“Still unsure about being vaccinated? Let’s talk about it.”

We aptly called the area the “No Judgment Zone.”

The No Judgment Zone is collaboratively staffed in 1- to 2-hour voluntary increments by physician faculty and resident physicians at Emory University School of Medicine and Morehouse School of Medicine. Our goal is to increase patient trust by honoring individual vaccine-related concerns without shame or ridicule. We also work to increase patient trust by being transparent around our own experiences with COVID-19; by sharing our own journeys, concerns, and challenges, we are better able to engage in meaningful dialogue. Also, recognizing the power of logistical barriers, in addition to answering questions, we offer physical assistance with check-in, forms, and escorts to our administration areas. The numbers of unique visits have varied from day to day, but the impact of each individual encounter cannot be overstated.

Here, we describe our approach to interactions at the No Judgment Zone. These are the instructions offered to our volunteers. Though we offer some explicit examples, each talking point is designed to open the door to a patient-centered individual dialogue. We believe that these strategies can be applied to clinical settings as well as any conversation surrounding vaccination with those who have not yet decided to be vaccinated.

No Labels

Try to think of all who are not yet vaccinated as “on a spectrum of deliberation” about their decision—not “hesitant” or “anti-vaxxer.”

Step 1: Gratitude

• “Thank you for stopping to talk to us today.”
• “I appreciate you taking the time.”
• “Before we start—I’m glad you’re here. Thanks.”

Step 2: Determine Where They Are

• Has the person you’re speaking with been vaccinated yet?
• If no, ask: “On a scale of 0 to 10—zero being “I will never get vaccinated under any circumstances” and 10 being ‘I will definitely get vaccinated’—what number would you give yourself?”
• If the person is a firm zero: “Is there anything I might be able to share with you or tell you about that might move you away from that perspective?”
• If the answer is NO: “It sounds like you’ve thought a lot about this and are no longer deliberating about whether you will be vaccinated. If you find yourself considering it, come back to talk with us, okay?” We are not here to debate or argue. We also need to avail ourselves to those who are open to changing their mind.
• If they say anything other than zero, move to an open-ended question about #WhatsYourWhy.

Step 3: #WhatsYourWhy

• “What would you say has been your main reason for not getting vaccinated yet?”
• “Tell me what has stood in the way of you getting vaccinated.”
• Remember: Assume nothing. It may have nothing to do with misinformation, fear, or perceived threat. It could be logistics or many other things. You will not know unless you ask.
• Providers should feel encouraged to also share their why as well and the reasons they encouraged their parents/kids/loved ones to get vaccinated. Making it personal can help establish connection and be more compelling.

Step 4: Listen Completely

• Give full eye contact. Slow all body movements. Use facilitative gestures to let the person know you are listening.
• Do not plan what you wish to say next.
• Limit reactions to misinformation. Shame and judgment can be subtle. Be mindful.
• Repeat the concern back if you are not sure or want to confirm that you’ve heard correctly.
• Ask questions for clarity if you aren’t sure.

Step 5: Affirm All Concerns and Find Common Ground

• “I can only imagine how scary it must be to take a shot that you believe could cause you to not be able to have babies.”
• “You aren’t alone. That’s a concern that many of my patients have had, too. May I share some information about that with you?”
• “When I first heard about the vaccine, I worried it was too new, too. Can I share what I learned?”

Step 6: Provide Factual Information

• Without excessive medical jargon, offer factual information aimed at each concern or question. Probe to be certain your patient understands through a teach-back or question.
• If you are unsure about the answer to their question, admit that you don’t know. You can also ask a colleague or the attending with you. Another option is to call someone or say “Let’s pull this up together.” Then share your answer.
• It is okay to acknowledge that the healthcare system has not and does not always do right by minority populations, especially Black people. Use that as a pivot to why these truths make vaccination that much more important
• Have FAQ information sheets available. Confirm that the patient is comfortable with the information sheet by asking.

Step 7: Offer to Help Them Get Vaccinated Today

• “Would you like me to help you get vaccinated today?”
• “What can I do to assist you with getting vaccinated? Is today a good day?”
• Escort those who agree to the registration area.
• Affirm those plans to get vaccinated or those who feel closer to getting vaccinated after speaking with you.

Step 8: Gratitude

• Close with gratitude and an affirmation.
• “I’m so glad you took the time to talk with us today. You didn’t have to stop.”
• “Feel free to come back to talk to us if you think of any more questions. I’m grateful that you stopped.”
• We are planting seeds. Do not feel pressure to get a person to say yes. Our secret sauce is kindness, respect, and empathy.
• We do not think of our unvaccinated community members as “hesitant.” We approach all as if they are on a spectrum of deliberation.

Step 9: Reflect

• Understand the importance of your service and the potential impact each encounter has.
• Recognize the unique lived experiences of individual patients and how this may impact their deliberation process. While there is urgency and we may feel frustrated, the ultimate goal is to engender trust through respectful interactions.
• Pause for moments of quiet gratitude and self-check-ins.

Just as SARS-CoV2 spreads from one person to many, we recognize that information—factual and otherwise—has the potential to move quickly as well. It is important to realize that providing an opportunity for people to ask questions or receive clarification and confirmation in a safe space is critical. The No Judgement Zone, as the name indicates, offers this opportunity. The conversations that we have in this space are valuable to those who are still considering the vaccine as an option for themselves. The trust required for such conversations is less about the transmission of information and more about the social act of engaging in bidirectional dialogue. The foundation upon which trust is built is consistent trustworthy actions. One such action is respectful communication without shame or ridicule. Another is our willingness to be transparent about our own concerns, experiences, and journeys. Assumptions based upon single-story narratives of the unvaccinated—particularly those from historically marginalized groups—fracture an already fragile confidence in medical authorities.

While we understand that mitigating the ongoing spread of the virus and getting more people vaccinated will call for more than just individual conversations, we believe that respecting the unique perspectives of community members is an equally critical piece to moving forward. Throughout a healthcare worker’s typical day, we work to create personal moments of connection with patients among the immense bustle of other work that has to be done. Initiatives like this one have a focused intentionality behind creating space for patients to feel heard that is not only helpful for vaccine uptake and addressing mistrust, but can also be restorative for providers as well.

The collective struggle felt by healthcare workers simultaneously learning about and caring for patients impacted by SARS-CoV2 infections throughout 2020 was physically and emotionally exhausting. The majority of us had never experienced a global pandemic. Beyond our work in the professional arena of ambulatory practices and hospitals, we also felt the soul-crushing impact of the pandemic in every other aspect of our lives. Preexisting health disparities were amplified by COVID-19. Some of the most affected communities also bore the weight of an additional tsunami of ongoing racial injustice.¹ And as healthcare workers, we did our best to process and navigate it all while trying to avoid burnout—as well as being infected with COVID-19 ourselves. When the news of the highly effective vaccines against SARS-CoV2 receiving emergency use authorization broke late in 2020, it felt like a light at the end of a very dark tunnel.

In the weeks preceding wide availability of the vaccines, it became apparent that significant numbers of people lacked confidence in the vaccines. Given the disproportionate impact of COVID-19 on racial minorities, much of the discussion centered around “vaccine hesitancy” in these communities. Reasons such as historical mistrust, belief in conspiracy theories, and misinformation emerged as the leading explanations.² Campaigns and educational programs targeting Black Americans were quickly developed to counter this widely distributed narrative.

Vaccine uptake also became politicized, which created additional challenges. As schools and businesses reopened, the voices of those opposing pandemic mitigation mandates such as masking and vaccination were highlighted by media outlets. And though a large movement of individuals who had opted against vaccines existed well before the pandemic, with few exceptions, that number had never been great enough to impact public health to this extent.³ This primarily nonminority group of unvaccinated individuals also morphed into another monolithic identity: the “anti-vaxxer.”

The lion’s share of discussions around vaccine uptake centered on these two groups: the “vaccine hesitant” minority and the “anti-vaxxer.” The perspectives and frustration around these two stereotypical unvaccinated groups were underscored in journals and the lay press. But those working in communities and in direct care came into contact with countless COVID-19-positive patients who were unvaccinated and fell into neither of these categories. There was a large swath of vulnerable people who still had unanswered questions and mistrust in the medical system standing in their way. Awareness of health disparities among racial minorities is something that was discussed among providers, but it was something experienced and felt by patients daily in regard to so much more than just COVID-19.

With broader access to vaccines through retail, community-based, and clinical facilities, more patients who desired vaccination had the opportunity. After an initial rise in vaccine uptake, the numbers plateaued. But what remained was the repetitive messaging and sustained focus directed toward Black people and their “vaccine hesitancy.”

Grady Memorial Hospital, a public safety net hospital in Atlanta, serves a predominantly Black and uninsured patient population. We found that a “FAQ” approach with a narrow range of hypothetical ideas about unvaccinated minorities clashed with the reality of what we encountered in clinical environments and the community. While misinformation did appear to be prevalent, we appreciated that the context and level of conviction were heterogenous. We appreciated that each individual conversation could reveal something new to us about that unique patient and their personal concerns about vaccination. As time moved forward, it became clear that there was no playbook for any group, especially for historically disadvantaged communities. Importantly, it was recognized that attempts to anticipate what may be a person’s barrier to vaccination often worked to further erode trust. However, when we focused on creating a space for dialogue, we found we were able to move beyond information-sharing and instead were able to co-construct interpretations of information and co-create solutions that matched patients’ values and lived experiences.⁴ Through dialogue, we were better able to be transparent about our own experiences, which ultimately facilitated authentic conversations with patients.

In September 2021, we approached our hospital leadership with a patient-centered strategy aimed at providing our patients, staff, and visitors a psychologically safe place to discuss vaccine-related concerns without judgment. With their support, we set up a table in the busiest part of our hospital atrium between the information desk and vaccine-administration site. Beside it was a folding board sign with an image and these words:

“Still unsure about being vaccinated? Let’s talk about it.”

We aptly called the area the “No Judgment Zone.”

The No Judgment Zone is collaboratively staffed in 1- to 2-hour voluntary increments by physician faculty and resident physicians at Emory University School of Medicine and Morehouse School of Medicine. Our goal is to increase patient trust by honoring individual vaccine-related concerns without shame or ridicule. We also work to increase patient trust by being transparent around our own experiences with COVID-19; by sharing our own journeys, concerns, and challenges, we are better able to engage in meaningful dialogue. Also, recognizing the power of logistical barriers, in addition to answering questions, we offer physical assistance with check-in, forms, and escorts to our administration areas. The numbers of unique visits have varied from day to day, but the impact of each individual encounter cannot be overstated.

Here, we describe our approach to interactions at the No Judgment Zone. These are the instructions offered to our volunteers. Though we offer some explicit examples, each talking point is designed to open the door to a patient-centered individual dialogue. We believe that these strategies can be applied to clinical settings as well as any conversation surrounding vaccination with those who have not yet decided to be vaccinated.

No Labels

Try to think of all who are not yet vaccinated as “on a spectrum of deliberation” about their decision—not “hesitant” or “anti-vaxxer.”

Step 1: Gratitude

• “Thank you for stopping to talk to us today.”
• “I appreciate you taking the time.”
• “Before we start—I’m glad you’re here. Thanks.”

Step 2: Determine Where They Are

• Has the person you’re speaking with been vaccinated yet?
• If no, ask: “On a scale of 0 to 10—zero being “I will never get vaccinated under any circumstances” and 10 being ‘I will definitely get vaccinated’—what number would you give yourself?”
• If the person is a firm zero: “Is there anything I might be able to share with you or tell you about that might move you away from that perspective?”
• If the answer is NO: “It sounds like you’ve thought a lot about this and are no longer deliberating about whether you will be vaccinated. If you find yourself considering it, come back to talk with us, okay?” We are not here to debate or argue. We also need to avail ourselves to those who are open to changing their mind.
• If they say anything other than zero, move to an open-ended question about #WhatsYourWhy.

Step 3: #WhatsYourWhy

• “What would you say has been your main reason for not getting vaccinated yet?”
• “Tell me what has stood in the way of you getting vaccinated.”
• Remember: Assume nothing. It may have nothing to do with misinformation, fear, or perceived threat. It could be logistics or many other things. You will not know unless you ask.
• Providers should feel encouraged to also share their why as well and the reasons they encouraged their parents/kids/loved ones to get vaccinated. Making it personal can help establish connection and be more compelling.

Step 4: Listen Completely

• Give full eye contact. Slow all body movements. Use facilitative gestures to let the person know you are listening.
• Do not plan what you wish to say next.
• Limit reactions to misinformation. Shame and judgment can be subtle. Be mindful.
• Repeat the concern back if you are not sure or want to confirm that you’ve heard correctly.
• Ask questions for clarity if you aren’t sure.

Step 5: Affirm All Concerns and Find Common Ground

• “I can only imagine how scary it must be to take a shot that you believe could cause you to not be able to have babies.”
• “You aren’t alone. That’s a concern that many of my patients have had, too. May I share some information about that with you?”
• “When I first heard about the vaccine, I worried it was too new, too. Can I share what I learned?”

Step 6: Provide Factual Information

• Without excessive medical jargon, offer factual information aimed at each concern or question. Probe to be certain your patient understands through a teach-back or question.
• If you are unsure about the answer to their question, admit that you don’t know. You can also ask a colleague or the attending with you. Another option is to call someone or say “Let’s pull this up together.” Then share your answer.
• It is okay to acknowledge that the healthcare system has not and does not always do right by minority populations, especially Black people. Use that as a pivot to why these truths make vaccination that much more important
• Have FAQ information sheets available. Confirm that the patient is comfortable with the information sheet by asking.

Step 7: Offer to Help Them Get Vaccinated Today

• “Would you like me to help you get vaccinated today?”
• “What can I do to assist you with getting vaccinated? Is today a good day?”
• Escort those who agree to the registration area.
• Affirm those plans to get vaccinated or those who feel closer to getting vaccinated after speaking with you.

Step 8: Gratitude

• Close with gratitude and an affirmation.
• “I’m so glad you took the time to talk with us today. You didn’t have to stop.”
• “Feel free to come back to talk to us if you think of any more questions. I’m grateful that you stopped.”
• We are planting seeds. Do not feel pressure to get a person to say yes. Our secret sauce is kindness, respect, and empathy.
• We do not think of our unvaccinated community members as “hesitant.” We approach all as if they are on a spectrum of deliberation.

Step 9: Reflect

• Understand the importance of your service and the potential impact each encounter has.
• Recognize the unique lived experiences of individual patients and how this may impact their deliberation process. While there is urgency and we may feel frustrated, the ultimate goal is to engender trust through respectful interactions.
• Pause for moments of quiet gratitude and self-check-ins.

Just as SARS-CoV2 spreads from one person to many, we recognize that information—factual and otherwise—has the potential to move quickly as well. It is important to realize that providing an opportunity for people to ask questions or receive clarification and confirmation in a safe space is critical. The No Judgement Zone, as the name indicates, offers this opportunity. The conversations that we have in this space are valuable to those who are still considering the vaccine as an option for themselves. The trust required for such conversations is less about the transmission of information and more about the social act of engaging in bidirectional dialogue. The foundation upon which trust is built is consistent trustworthy actions. One such action is respectful communication without shame or ridicule. Another is our willingness to be transparent about our own concerns, experiences, and journeys. Assumptions based upon single-story narratives of the unvaccinated—particularly those from historically marginalized groups—fracture an already fragile confidence in medical authorities.

While we understand that mitigating the ongoing spread of the virus and getting more people vaccinated will call for more than just individual conversations, we believe that respecting the unique perspectives of community members is an equally critical piece to moving forward. Throughout a healthcare worker’s typical day, we work to create personal moments of connection with patients among the immense bustle of other work that has to be done. Initiatives like this one have a focused intentionality behind creating space for patients to feel heard that is not only helpful for vaccine uptake and addressing mistrust, but can also be restorative for providers as well.

The collective struggle felt by healthcare workers simultaneously learning about and caring for patients impacted by SARS-CoV2 infections throughout 2020 was physically and emotionally exhausting. The majority of us had never experienced a global pandemic. Beyond our work in the professional arena of ambulatory practices and hospitals, we also felt the soul-crushing impact of the pandemic in every other aspect of our lives. Preexisting health disparities were amplified by COVID-19. Some of the most affected communities also bore the weight of an additional tsunami of ongoing racial injustice.¹ And as healthcare workers, we did our best to process and navigate it all while trying to avoid burnout—as well as being infected with COVID-19 ourselves. When the news of the highly effective vaccines against SARS-CoV2 receiving emergency use authorization broke late in 2020, it felt like a light at the end of a very dark tunnel.

In the weeks preceding wide availability of the vaccines, it became apparent that significant numbers of people lacked confidence in the vaccines. Given the disproportionate impact of COVID-19 on racial minorities, much of the discussion centered around “vaccine hesitancy” in these communities. Reasons such as historical mistrust, belief in conspiracy theories, and misinformation emerged as the leading explanations.² Campaigns and educational programs targeting Black Americans were quickly developed to counter this widely distributed narrative.

Vaccine uptake also became politicized, which created additional challenges. As schools and businesses reopened, the voices of those opposing pandemic mitigation mandates such as masking and vaccination were highlighted by media outlets. And though a large movement of individuals who had opted against vaccines existed well before the pandemic, with few exceptions, that number had never been great enough to impact public health to this extent.³ This primarily nonminority group of unvaccinated individuals also morphed into another monolithic identity: the “anti-vaxxer.”

The lion’s share of discussions around vaccine uptake centered on these two groups: the “vaccine hesitant” minority and the “anti-vaxxer.” The perspectives and frustration around these two stereotypical unvaccinated groups were underscored in journals and the lay press. But those working in communities and in direct care came into contact with countless COVID-19-positive patients who were unvaccinated and fell into neither of these categories. There was a large swath of vulnerable people who still had unanswered questions and mistrust in the medical system standing in their way. Awareness of health disparities among racial minorities is something that was discussed among providers, but it was something experienced and felt by patients daily in regard to so much more than just COVID-19.

With broader access to vaccines through retail, community-based, and clinical facilities, more patients who desired vaccination had the opportunity. After an initial rise in vaccine uptake, the numbers plateaued. But what remained was the repetitive messaging and sustained focus directed toward Black people and their “vaccine hesitancy.”

Grady Memorial Hospital, a public safety net hospital in Atlanta, serves a predominantly Black and uninsured patient population. We found that a “FAQ” approach with a narrow range of hypothetical ideas about unvaccinated minorities clashed with the reality of what we encountered in clinical environments and the community. While misinformation did appear to be prevalent, we appreciated that the context and level of conviction were heterogenous. We appreciated that each individual conversation could reveal something new to us about that unique patient and their personal concerns about vaccination. As time moved forward, it became clear that there was no playbook for any group, especially for historically disadvantaged communities. Importantly, it was recognized that attempts to anticipate what may be a person’s barrier to vaccination often worked to further erode trust. However, when we focused on creating a space for dialogue, we found we were able to move beyond information-sharing and instead were able to co-construct interpretations of information and co-create solutions that matched patients’ values and lived experiences.⁴ Through dialogue, we were better able to be transparent about our own experiences, which ultimately facilitated authentic conversations with patients.

In September 2021, we approached our hospital leadership with a patient-centered strategy aimed at providing our patients, staff, and visitors a psychologically safe place to discuss vaccine-related concerns without judgment. With their support, we set up a table in the busiest part of our hospital atrium between the information desk and vaccine-administration site. Beside it was a folding board sign with an image and these words:

“Still unsure about being vaccinated? Let’s talk about it.”

We aptly called the area the “No Judgment Zone.”

The No Judgment Zone is collaboratively staffed in 1- to 2-hour voluntary increments by physician faculty and resident physicians at Emory University School of Medicine and Morehouse School of Medicine. Our goal is to increase patient trust by honoring individual vaccine-related concerns without shame or ridicule. We also work to increase patient trust by being transparent around our own experiences with COVID-19; by sharing our own journeys, concerns, and challenges, we are better able to engage in meaningful dialogue. Also, recognizing the power of logistical barriers, in addition to answering questions, we offer physical assistance with check-in, forms, and escorts to our administration areas. The numbers of unique visits have varied from day to day, but the impact of each individual encounter cannot be overstated.

Here, we describe our approach to interactions at the No Judgment Zone. These are the instructions offered to our volunteers. Though we offer some explicit examples, each talking point is designed to open the door to a patient-centered individual dialogue. We believe that these strategies can be applied to clinical settings as well as any conversation surrounding vaccination with those who have not yet decided to be vaccinated.

No Labels

Try to think of all who are not yet vaccinated as “on a spectrum of deliberation” about their decision—not “hesitant” or “anti-vaxxer.”

Step 1: Gratitude

• “Thank you for stopping to talk to us today.”
• “I appreciate you taking the time.”
• “Before we start—I’m glad you’re here. Thanks.”

Step 2: Determine Where They Are

• Has the person you’re speaking with been vaccinated yet?
• If no, ask: “On a scale of 0 to 10—zero being “I will never get vaccinated under any circumstances” and 10 being ‘I will definitely get vaccinated’—what number would you give yourself?”
• If the person is a firm zero: “Is there anything I might be able to share with you or tell you about that might move you away from that perspective?”
• If the answer is NO: “It sounds like you’ve thought a lot about this and are no longer deliberating about whether you will be vaccinated. If you find yourself considering it, come back to talk with us, okay?” We are not here to debate or argue. We also need to avail ourselves to those who are open to changing their mind.
• If they say anything other than zero, move to an open-ended question about #WhatsYourWhy.

Step 3: #WhatsYourWhy

• “What would you say has been your main reason for not getting vaccinated yet?”
• “Tell me what has stood in the way of you getting vaccinated.”
• Remember: Assume nothing. It may have nothing to do with misinformation, fear, or perceived threat. It could be logistics or many other things. You will not know unless you ask.
• Providers should feel encouraged to also share their why as well and the reasons they encouraged their parents/kids/loved ones to get vaccinated. Making it personal can help establish connection and be more compelling.

Step 4: Listen Completely

• Give full eye contact. Slow all body movements. Use facilitative gestures to let the person know you are listening.
• Do not plan what you wish to say next.
• Limit reactions to misinformation. Shame and judgment can be subtle. Be mindful.
• Repeat the concern back if you are not sure or want to confirm that you’ve heard correctly.
• Ask questions for clarity if you aren’t sure.

Step 5: Affirm All Concerns and Find Common Ground

• “I can only imagine how scary it must be to take a shot that you believe could cause you to not be able to have babies.”
• “You aren’t alone. That’s a concern that many of my patients have had, too. May I share some information about that with you?”
• “When I first heard about the vaccine, I worried it was too new, too. Can I share what I learned?”

Step 6: Provide Factual Information

• Without excessive medical jargon, offer factual information aimed at each concern or question. Probe to be certain your patient understands through a teach-back or question.
• If you are unsure about the answer to their question, admit that you don’t know. You can also ask a colleague or the attending with you. Another option is to call someone or say “Let’s pull this up together.” Then share your answer.
• It is okay to acknowledge that the healthcare system has not and does not always do right by minority populations, especially Black people. Use that as a pivot to why these truths make vaccination that much more important
• Have FAQ information sheets available. Confirm that the patient is comfortable with the information sheet by asking.

Step 7: Offer to Help Them Get Vaccinated Today

• “Would you like me to help you get vaccinated today?”
• “What can I do to assist you with getting vaccinated? Is today a good day?”
• Escort those who agree to the registration area.
• Affirm those plans to get vaccinated or those who feel closer to getting vaccinated after speaking with you.

Step 8: Gratitude

• Close with gratitude and an affirmation.
• “I’m so glad you took the time to talk with us today. You didn’t have to stop.”
• “Feel free to come back to talk to us if you think of any more questions. I’m grateful that you stopped.”
• We are planting seeds. Do not feel pressure to get a person to say yes. Our secret sauce is kindness, respect, and empathy.
• We do not think of our unvaccinated community members as “hesitant.” We approach all as if they are on a spectrum of deliberation.

Step 9: Reflect

• Understand the importance of your service and the potential impact each encounter has.
• Recognize the unique lived experiences of individual patients and how this may impact their deliberation process. While there is urgency and we may feel frustrated, the ultimate goal is to engender trust through respectful interactions.
• Pause for moments of quiet gratitude and self-check-ins.

Just as SARS-CoV2 spreads from one person to many, we recognize that information—factual and otherwise—has the potential to move quickly as well. It is important to realize that providing an opportunity for people to ask questions or receive clarification and confirmation in a safe space is critical. The No Judgement Zone, as the name indicates, offers this opportunity. The conversations that we have in this space are valuable to those who are still considering the vaccine as an option for themselves. The trust required for such conversations is less about the transmission of information and more about the social act of engaging in bidirectional dialogue. The foundation upon which trust is built is consistent trustworthy actions. One such action is respectful communication without shame or ridicule. Another is our willingness to be transparent about our own concerns, experiences, and journeys. Assumptions based upon single-story narratives of the unvaccinated—particularly those from historically marginalized groups—fracture an already fragile confidence in medical authorities.

While we understand that mitigating the ongoing spread of the virus and getting more people vaccinated will call for more than just individual conversations, we believe that respecting the unique perspectives of community members is an equally critical piece to moving forward. Throughout a healthcare worker’s typical day, we work to create personal moments of connection with patients among the immense bustle of other work that has to be done. Initiatives like this one have a focused intentionality behind creating space for patients to feel heard that is not only helpful for vaccine uptake and addressing mistrust, but can also be restorative for providers as well.

Naila Makhani, MD completed medical school training at the University of British Columbia (Vancouver, Canada). This was followed by a residency in child neurology and fellowship in MS and other demyelinating diseases at the University of Toronto and The Hospital for Sick Children (Toronto, Canada). Concurrent with fellowship training, Dr. Makhani obtained a Masters’ degree in public health from Harvard University. Dr. Makhani is the Director of the Pediatric MS Program at Yale and the lead investigator of a multi-center international study examining outcomes following the radiologically isolated syndrome in children.

Q1. Could you please provide an overview of Radiologically Isolated Syndrome ?

A1. Radiologically Isolated Syndrome (RIS) was first described in adults in 2009. Since then it has also been increasingly recognized and diagnosed in children. RIS is diagnosed after an MRI of the brain that the patient has sought for reasons other than suspected multiple sclerosis-- for instance, for evaluation of head trauma or headache. However, unexpectedly or incidentally, the patient’s MRI shows the typical findings that we see in multiple sclerosis, even in the absence of any typical clinical symptoms. RIS is generally considered a rare syndrome.

Q2. You created Yale Medicine’s Pediatric Multiple Sclerosis program which advocates for the eradication of MS. What  criteria defines a rare disease? Does RIS meet these criteria? And if so, how?

A2. The criteria for a rare disease vary, depending on the reference. In the US, a rare disease is defined as a condition that affects fewer than 200,000 people, in total, across the country.  By contrast, in Europe, a disease is considered rare if it affects fewer than one in every 2,000 people within the country’s population.

In the case of RIS, especially in children, we suspect that this is a rare condition, but we don't know for sure, as there have been very few population-based studies. There is one large study that was conducted in Europe that found one case of RIS among approximately 5,000 otherwise healthy children, who were between 7 and 14 years of age.  I think that's our best estimate of the overall prevalence of RIS in children. Using that finding, it likely would qualify as a rare condition, although, as I said, we really don't know for sure, as the prevalence may vary among different populations or age groups.

Q3. How do you investigate and manage RIS in children? What are some of the challenges?  

A3. For children with radiologically isolated syndrome, we usually undertake a comprehensive workup. This includes a detailed clinical neurological exam to ensure that there are no abnormalities that would, for instance, suggest a misdiagnosis of multiple sclerosis or an alternative diagnosis. In addition to the brain MRI, we usually obtain an MRI of the spinal cord to determine whether there is any spinal cord involvement. We also obtain blood tests. We often analyze spinal fluid as well, primarily to exclude other alternative processes that may explain the  MRI findings. A key challenge in this field is that there are currently no formal guidelines for the investigation and management of children with RIS. Collaborations within the pediatric MS community are needed to develop such consensus approaches to standardize care.

Q4. What are the most significant risk factors that indicate children with RIS could one day develop multiple sclerosis?

A4.This is an area of active research within our group. So far, we've found that approximately 42% of children with RIS develop multiple sclerosis in the future; on average, two years following their first abnormal MRI. Therefore, this is a high-risk group for developing  multiple sclerosis in the future. Thus far, we've determined that in children with RIS, it is the presence of abnormal spinal cord imaging and an abnormality in spinal fluid – namely, the presence of oligoclonal bands – that are likely the predictors of whether these children could develop MS in the future. a child’s possible development 

Q5. Based on your recent studies, are there data in children highlighting the potential for higher prevalence  in one population over another?

A5. Thus far, population-based studies assessing RIS, especially in children, have been rare and thus far have not identified particular subgroups with increased prevalence. We do know that the prevalence of multiple sclerosis varies across different age groups and across gender. Whether such associations are also present for RIS is an area of active research.

Naila Makhani, MD completed medical school training at the University of British Columbia (Vancouver, Canada). This was followed by a residency in child neurology and fellowship in MS and other demyelinating diseases at the University of Toronto and The Hospital for Sick Children (Toronto, Canada). Concurrent with fellowship training, Dr. Makhani obtained a Masters’ degree in public health from Harvard University. Dr. Makhani is the Director of the Pediatric MS Program at Yale and the lead investigator of a multi-center international study examining outcomes following the radiologically isolated syndrome in children.

Q1. Could you please provide an overview of Radiologically Isolated Syndrome ?

A1. Radiologically Isolated Syndrome (RIS) was first described in adults in 2009. Since then it has also been increasingly recognized and diagnosed in children. RIS is diagnosed after an MRI of the brain that the patient has sought for reasons other than suspected multiple sclerosis-- for instance, for evaluation of head trauma or headache. However, unexpectedly or incidentally, the patient’s MRI shows the typical findings that we see in multiple sclerosis, even in the absence of any typical clinical symptoms. RIS is generally considered a rare syndrome.

Q2. You created Yale Medicine’s Pediatric Multiple Sclerosis program which advocates for the eradication of MS. What  criteria defines a rare disease? Does RIS meet these criteria? And if so, how?

A2. The criteria for a rare disease vary, depending on the reference. In the US, a rare disease is defined as a condition that affects fewer than 200,000 people, in total, across the country.  By contrast, in Europe, a disease is considered rare if it affects fewer than one in every 2,000 people within the country’s population.

In the case of RIS, especially in children, we suspect that this is a rare condition, but we don't know for sure, as there have been very few population-based studies. There is one large study that was conducted in Europe that found one case of RIS among approximately 5,000 otherwise healthy children, who were between 7 and 14 years of age.  I think that's our best estimate of the overall prevalence of RIS in children. Using that finding, it likely would qualify as a rare condition, although, as I said, we really don't know for sure, as the prevalence may vary among different populations or age groups.

Q3. How do you investigate and manage RIS in children? What are some of the challenges?  

A3. For children with radiologically isolated syndrome, we usually undertake a comprehensive workup. This includes a detailed clinical neurological exam to ensure that there are no abnormalities that would, for instance, suggest a misdiagnosis of multiple sclerosis or an alternative diagnosis. In addition to the brain MRI, we usually obtain an MRI of the spinal cord to determine whether there is any spinal cord involvement. We also obtain blood tests. We often analyze spinal fluid as well, primarily to exclude other alternative processes that may explain the  MRI findings. A key challenge in this field is that there are currently no formal guidelines for the investigation and management of children with RIS. Collaborations within the pediatric MS community are needed to develop such consensus approaches to standardize care.

Q4. What are the most significant risk factors that indicate children with RIS could one day develop multiple sclerosis?

A4.This is an area of active research within our group. So far, we've found that approximately 42% of children with RIS develop multiple sclerosis in the future; on average, two years following their first abnormal MRI. Therefore, this is a high-risk group for developing  multiple sclerosis in the future. Thus far, we've determined that in children with RIS, it is the presence of abnormal spinal cord imaging and an abnormality in spinal fluid – namely, the presence of oligoclonal bands – that are likely the predictors of whether these children could develop MS in the future. a child’s possible development 

Q5. Based on your recent studies, are there data in children highlighting the potential for higher prevalence  in one population over another?

A5. Thus far, population-based studies assessing RIS, especially in children, have been rare and thus far have not identified particular subgroups with increased prevalence. We do know that the prevalence of multiple sclerosis varies across different age groups and across gender. Whether such associations are also present for RIS is an area of active research.

Naila Makhani, MD completed medical school training at the University of British Columbia (Vancouver, Canada). This was followed by a residency in child neurology and fellowship in MS and other demyelinating diseases at the University of Toronto and The Hospital for Sick Children (Toronto, Canada). Concurrent with fellowship training, Dr. Makhani obtained a Masters’ degree in public health from Harvard University. Dr. Makhani is the Director of the Pediatric MS Program at Yale and the lead investigator of a multi-center international study examining outcomes following the radiologically isolated syndrome in children.

Q1. Could you please provide an overview of Radiologically Isolated Syndrome ?

A1. Radiologically Isolated Syndrome (RIS) was first described in adults in 2009. Since then it has also been increasingly recognized and diagnosed in children. RIS is diagnosed after an MRI of the brain that the patient has sought for reasons other than suspected multiple sclerosis-- for instance, for evaluation of head trauma or headache. However, unexpectedly or incidentally, the patient’s MRI shows the typical findings that we see in multiple sclerosis, even in the absence of any typical clinical symptoms. RIS is generally considered a rare syndrome.

Q2. You created Yale Medicine’s Pediatric Multiple Sclerosis program which advocates for the eradication of MS. What  criteria defines a rare disease? Does RIS meet these criteria? And if so, how?

A2. The criteria for a rare disease vary, depending on the reference. In the US, a rare disease is defined as a condition that affects fewer than 200,000 people, in total, across the country.  By contrast, in Europe, a disease is considered rare if it affects fewer than one in every 2,000 people within the country’s population.

In the case of RIS, especially in children, we suspect that this is a rare condition, but we don't know for sure, as there have been very few population-based studies. There is one large study that was conducted in Europe that found one case of RIS among approximately 5,000 otherwise healthy children, who were between 7 and 14 years of age.  I think that's our best estimate of the overall prevalence of RIS in children. Using that finding, it likely would qualify as a rare condition, although, as I said, we really don't know for sure, as the prevalence may vary among different populations or age groups.

Q3. How do you investigate and manage RIS in children? What are some of the challenges?  

A3. For children with radiologically isolated syndrome, we usually undertake a comprehensive workup. This includes a detailed clinical neurological exam to ensure that there are no abnormalities that would, for instance, suggest a misdiagnosis of multiple sclerosis or an alternative diagnosis. In addition to the brain MRI, we usually obtain an MRI of the spinal cord to determine whether there is any spinal cord involvement. We also obtain blood tests. We often analyze spinal fluid as well, primarily to exclude other alternative processes that may explain the  MRI findings. A key challenge in this field is that there are currently no formal guidelines for the investigation and management of children with RIS. Collaborations within the pediatric MS community are needed to develop such consensus approaches to standardize care.

Q4. What are the most significant risk factors that indicate children with RIS could one day develop multiple sclerosis?

A4.This is an area of active research within our group. So far, we've found that approximately 42% of children with RIS develop multiple sclerosis in the future; on average, two years following their first abnormal MRI. Therefore, this is a high-risk group for developing  multiple sclerosis in the future. Thus far, we've determined that in children with RIS, it is the presence of abnormal spinal cord imaging and an abnormality in spinal fluid – namely, the presence of oligoclonal bands – that are likely the predictors of whether these children could develop MS in the future. a child’s possible development 

Q5. Based on your recent studies, are there data in children highlighting the potential for higher prevalence  in one population over another?

A5. Thus far, population-based studies assessing RIS, especially in children, have been rare and thus far have not identified particular subgroups with increased prevalence. We do know that the prevalence of multiple sclerosis varies across different age groups and across gender. Whether such associations are also present for RIS is an area of active research.

Naila Makhani, MD completed medical school training at the University of British Columbia (Vancouver, Canada). This was followed by a residency in child neurology and fellowship in MS and other demyelinating diseases at the University of Toronto and The Hospital for Sick Children (Toronto, Canada). Concurrent with fellowship training, Dr. Makhani obtained a Masters’ degree in public health from Harvard University. Dr. Makhani is the Director of the Pediatric MS Program at Yale and the lead investigator of a multi-center international study examining outcomes following the radiologically isolated syndrome in children.

Q1. Could you please provide an overview of Radiologically Isolated Syndrome ?

A1. Radiologically Isolated Syndrome (RIS) was first described in adults in 2009. Since then it has also been increasingly recognized and diagnosed in children. RIS is diagnosed after an MRI of the brain that the patient has sought for reasons other than suspected multiple sclerosis-- for instance, for evaluation of head trauma or headache. However, unexpectedly or incidentally, the patient’s MRI shows the typical findings that we see in multiple sclerosis, even in the absence of any typical clinical symptoms. RIS is generally considered a rare syndrome.

Q2. You created Yale Medicine’s Pediatric Multiple Sclerosis program which advocates for the eradication of MS. What  criteria defines a rare disease? Does RIS meet these criteria? And if so, how?

A2. The criteria for a rare disease vary, depending on the reference. In the US, a rare disease is defined as a condition that affects fewer than 200,000 people, in total, across the country.  By contrast, in Europe, a disease is considered rare if it affects fewer than one in every 2,000 people within the country’s population.

In the case of RIS, especially in children, we suspect that this is a rare condition, but we don't know for sure, as there have been very few population-based studies. There is one large study that was conducted in Europe that found one case of RIS among approximately 5,000 otherwise healthy children, who were between 7 and 14 years of age.  I think that's our best estimate of the overall prevalence of RIS in children. Using that finding, it likely would qualify as a rare condition, although, as I said, we really don't know for sure, as the prevalence may vary among different populations or age groups.

Q3. How do you investigate and manage RIS in children? What are some of the challenges?  

A3. For children with radiologically isolated syndrome, we usually undertake a comprehensive workup. This includes a detailed clinical neurological exam to ensure that there are no abnormalities that would, for instance, suggest a misdiagnosis of multiple sclerosis or an alternative diagnosis. In addition to the brain MRI, we usually obtain an MRI of the spinal cord to determine whether there is any spinal cord involvement. We also obtain blood tests. We often analyze spinal fluid as well, primarily to exclude other alternative processes that may explain the  MRI findings. A key challenge in this field is that there are currently no formal guidelines for the investigation and management of children with RIS. Collaborations within the pediatric MS community are needed to develop such consensus approaches to standardize care.

Q4. What are the most significant risk factors that indicate children with RIS could one day develop multiple sclerosis?

A4.This is an area of active research within our group. So far, we've found that approximately 42% of children with RIS develop multiple sclerosis in the future; on average, two years following their first abnormal MRI. Therefore, this is a high-risk group for developing  multiple sclerosis in the future. Thus far, we've determined that in children with RIS, it is the presence of abnormal spinal cord imaging and an abnormality in spinal fluid – namely, the presence of oligoclonal bands – that are likely the predictors of whether these children could develop MS in the future. a child’s possible development 

Q5. Based on your recent studies, are there data in children highlighting the potential for higher prevalence  in one population over another?

A5. Thus far, population-based studies assessing RIS, especially in children, have been rare and thus far have not identified particular subgroups with increased prevalence. We do know that the prevalence of multiple sclerosis varies across different age groups and across gender. Whether such associations are also present for RIS is an area of active research.

Naila Makhani, MD completed medical school training at the University of British Columbia (Vancouver, Canada). This was followed by a residency in child neurology and fellowship in MS and other demyelinating diseases at the University of Toronto and The Hospital for Sick Children (Toronto, Canada). Concurrent with fellowship training, Dr. Makhani obtained a Masters’ degree in public health from Harvard University. Dr. Makhani is the Director of the Pediatric MS Program at Yale and the lead investigator of a multi-center international study examining outcomes following the radiologically isolated syndrome in children.

Q1. Could you please provide an overview of Radiologically Isolated Syndrome ?

A1. Radiologically Isolated Syndrome (RIS) was first described in adults in 2009. Since then it has also been increasingly recognized and diagnosed in children. RIS is diagnosed after an MRI of the brain that the patient has sought for reasons other than suspected multiple sclerosis-- for instance, for evaluation of head trauma or headache. However, unexpectedly or incidentally, the patient’s MRI shows the typical findings that we see in multiple sclerosis, even in the absence of any typical clinical symptoms. RIS is generally considered a rare syndrome.

Q2. You created Yale Medicine’s Pediatric Multiple Sclerosis program which advocates for the eradication of MS. What  criteria defines a rare disease? Does RIS meet these criteria? And if so, how?

A2. The criteria for a rare disease vary, depending on the reference. In the US, a rare disease is defined as a condition that affects fewer than 200,000 people, in total, across the country.  By contrast, in Europe, a disease is considered rare if it affects fewer than one in every 2,000 people within the country’s population.

In the case of RIS, especially in children, we suspect that this is a rare condition, but we don't know for sure, as there have been very few population-based studies. There is one large study that was conducted in Europe that found one case of RIS among approximately 5,000 otherwise healthy children, who were between 7 and 14 years of age.  I think that's our best estimate of the overall prevalence of RIS in children. Using that finding, it likely would qualify as a rare condition, although, as I said, we really don't know for sure, as the prevalence may vary among different populations or age groups.

Q3. How do you investigate and manage RIS in children? What are some of the challenges?  

A3. For children with radiologically isolated syndrome, we usually undertake a comprehensive workup. This includes a detailed clinical neurological exam to ensure that there are no abnormalities that would, for instance, suggest a misdiagnosis of multiple sclerosis or an alternative diagnosis. In addition to the brain MRI, we usually obtain an MRI of the spinal cord to determine whether there is any spinal cord involvement. We also obtain blood tests. We often analyze spinal fluid as well, primarily to exclude other alternative processes that may explain the  MRI findings. A key challenge in this field is that there are currently no formal guidelines for the investigation and management of children with RIS. Collaborations within the pediatric MS community are needed to develop such consensus approaches to standardize care.

Q4. What are the most significant risk factors that indicate children with RIS could one day develop multiple sclerosis?

A4.This is an area of active research within our group. So far, we've found that approximately 42% of children with RIS develop multiple sclerosis in the future; on average, two years following their first abnormal MRI. Therefore, this is a high-risk group for developing  multiple sclerosis in the future. Thus far, we've determined that in children with RIS, it is the presence of abnormal spinal cord imaging and an abnormality in spinal fluid – namely, the presence of oligoclonal bands – that are likely the predictors of whether these children could develop MS in the future. a child’s possible development 

Q5. Based on your recent studies, are there data in children highlighting the potential for higher prevalence  in one population over another?

A5. Thus far, population-based studies assessing RIS, especially in children, have been rare and thus far have not identified particular subgroups with increased prevalence. We do know that the prevalence of multiple sclerosis varies across different age groups and across gender. Whether such associations are also present for RIS is an area of active research.

Naila Makhani, MD completed medical school training at the University of British Columbia (Vancouver, Canada). This was followed by a residency in child neurology and fellowship in MS and other demyelinating diseases at the University of Toronto and The Hospital for Sick Children (Toronto, Canada). Concurrent with fellowship training, Dr. Makhani obtained a Masters’ degree in public health from Harvard University. Dr. Makhani is the Director of the Pediatric MS Program at Yale and the lead investigator of a multi-center international study examining outcomes following the radiologically isolated syndrome in children.

Q1. Could you please provide an overview of Radiologically Isolated Syndrome ?

A1. Radiologically Isolated Syndrome (RIS) was first described in adults in 2009. Since then it has also been increasingly recognized and diagnosed in children. RIS is diagnosed after an MRI of the brain that the patient has sought for reasons other than suspected multiple sclerosis-- for instance, for evaluation of head trauma or headache. However, unexpectedly or incidentally, the patient’s MRI shows the typical findings that we see in multiple sclerosis, even in the absence of any typical clinical symptoms. RIS is generally considered a rare syndrome.

Q2. You created Yale Medicine’s Pediatric Multiple Sclerosis program which advocates for the eradication of MS. What  criteria defines a rare disease? Does RIS meet these criteria? And if so, how?

A2. The criteria for a rare disease vary, depending on the reference. In the US, a rare disease is defined as a condition that affects fewer than 200,000 people, in total, across the country.  By contrast, in Europe, a disease is considered rare if it affects fewer than one in every 2,000 people within the country’s population.

In the case of RIS, especially in children, we suspect that this is a rare condition, but we don't know for sure, as there have been very few population-based studies. There is one large study that was conducted in Europe that found one case of RIS among approximately 5,000 otherwise healthy children, who were between 7 and 14 years of age.  I think that's our best estimate of the overall prevalence of RIS in children. Using that finding, it likely would qualify as a rare condition, although, as I said, we really don't know for sure, as the prevalence may vary among different populations or age groups.

Q3. How do you investigate and manage RIS in children? What are some of the challenges?  

A3. For children with radiologically isolated syndrome, we usually undertake a comprehensive workup. This includes a detailed clinical neurological exam to ensure that there are no abnormalities that would, for instance, suggest a misdiagnosis of multiple sclerosis or an alternative diagnosis. In addition to the brain MRI, we usually obtain an MRI of the spinal cord to determine whether there is any spinal cord involvement. We also obtain blood tests. We often analyze spinal fluid as well, primarily to exclude other alternative processes that may explain the  MRI findings. A key challenge in this field is that there are currently no formal guidelines for the investigation and management of children with RIS. Collaborations within the pediatric MS community are needed to develop such consensus approaches to standardize care.

Q4. What are the most significant risk factors that indicate children with RIS could one day develop multiple sclerosis?

A4.This is an area of active research within our group. So far, we've found that approximately 42% of children with RIS develop multiple sclerosis in the future; on average, two years following their first abnormal MRI. Therefore, this is a high-risk group for developing  multiple sclerosis in the future. Thus far, we've determined that in children with RIS, it is the presence of abnormal spinal cord imaging and an abnormality in spinal fluid – namely, the presence of oligoclonal bands – that are likely the predictors of whether these children could develop MS in the future. a child’s possible development 

Q5. Based on your recent studies, are there data in children highlighting the potential for higher prevalence  in one population over another?

A5. Thus far, population-based studies assessing RIS, especially in children, have been rare and thus far have not identified particular subgroups with increased prevalence. We do know that the prevalence of multiple sclerosis varies across different age groups and across gender. Whether such associations are also present for RIS is an area of active research.

For those of us who cannot sit in the sun and fish all day, the next best thing for preventing autoimmune diseases may be supplementation with vitamin D and fish oil-derived omega-3 fatty acids, results of a large prospective randomized trial suggest.

Ziga Plahutar

Among nearly 26,000 adults enrolled in a randomized trial designed primarily to study the effects of vitamin D and omega-3 supplementation on incident cancer and cardiovascular disease, 5 years of vitamin D supplementation was associated with a 22% reduction in risk for confirmed autoimmune diseases, and 5 years of omega-3 fatty acid supplementation was associated with an 18% reduction in confirmed and probable incident autoimmune diseases, reported Karen H. Costenbader, MD, MPH, of Brigham & Women’s Hospital in Boston.

“The clinical importance of these results is very high, given that these are nontoxic, well-tolerated supplements, and that there are no other known effective therapies to reduce the incidence of autoimmune diseases,” she said during the virtual annual meeting of the American College of Rheumatology.

“People do have to take the supplements a long time to start to see the reduction in risk, especially for vitamin D, but they make biological sense, and autoimmune diseases develop slowly over time, so taking it today isn’t going to reduce risk of developing something tomorrow,” Dr. Costenbader said in an interview.

“These supplements have other health benefits. Obviously, fish oil is anti-inflammatory, and vitamin D is good for osteoporosis prevention, especially in our patients who take glucocorticoids. People who are otherwise healthy and have a family history of autoimmune disease might also consider starting to take these supplements,” she said.

After watching her presentation, session co-moderator Gregg Silverman, MD, from the NYU Langone School of Medicine in New York, who was not involved in the study, commented “I’m going to [nutrition store] GNC to get some vitamins.”

When asked for comment, the other session moderator, Tracy Frech, MD, of Vanderbilt University, Nashville, said, “I think Dr. Costenbader’s work is very important and her presentation excellent. My current practice is replacement of vitamin D in all autoimmune disease patients with low levels and per bone health guidelines. Additionally, I discuss omega-3 supplementation with Sjögren’s [syndrome] patients as a consideration.”

Evidence base

Dr. Costenbader noted that in a 2013 observational study from France, vitamin D derived through ultraviolet (UV) light exposure was associated with a lower risk for incident Crohn’s disease but not ulcerative colitis, and in two analyses of data in 2014 from the Nurses’ Health Study, both high plasma levels of 25-OH vitamin D and geographic residence in areas of high UV exposure were associated with a decreased incidence of rheumatoid arthritis (RA).

Other observational studies have supported omega-3 fatty acids for their anti-inflammatory properties, including a 2005 Danish prospective cohort study showing a lower risk for RA in participants who reported higher levels of fatty fish intake. In a separate study conducted in 2017, healthy volunteers with higher omega-3 fatty acid/total lipid proportions in red blood cell membranes had a lower prevalence of anti-cyclic citrullinated peptide (anti-CCP) antibodies and rheumatoid factor and a lower incidence of progression to inflammatory arthritis, she said.

Ancillary study

Despite the evidence, however, there have been no prospective randomized trials to test the effects of either vitamin D or omega-3 fatty acid supplementation on the incidence of autoimmune disease over time.

To rectify this, Dr. Costenbader and colleagues piggybacked an ancillary study onto the Vitamin D and Omega-3 Trial (VITAL), which had primary outcomes of cancer and cardiovascular disease incidence.

A total of 25,871 participants were enrolled, including 12,786 men aged 50 and older, and 13,085 women aged 55 and older.

The study had a 2 x 2 factorial design, with patients randomly assigned to vitamin D 2,000 IU/day or placebo, and then further randomized to either 1 g/day omega-3 fatty acids or placebo in both the vitamin D and placebo primary randomization arms.

At baseline 16,956 participants were assayed for 25-OH vitamin D and plasma omega 3 index, the ratio of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) to total fatty acids. Participants self-reported baseline and all incident autoimmune diseases annually, with the reports confirmed by medical record review and disease criteria whenever possible.

Results

At 5 years of follow-up, confirmed incident autoimmune diseases had occurred in 123 patients in the active vitamin D group, compared with 155 in the placebo vitamin D group, translating into a hazard ratio (HR) for vitamin D of 0.78 (P = .045).

In the active omega-3 arm, 130 participants developed an autoimmune disease, compared with 148 in the placebo omega-3 arm, which translated into a nonsignificant HR of 0.85.

There was no statistical interaction between the two supplements. The investigators did observe an interaction between vitamin D and body mass index, with the effect stronger among participants with low BMI (P = .02). There also was an interaction between omega-3 fatty acids with a family history of autoimmune disease (P = .03).

In multivariate analysis adjusted for age, sex, race, and other supplement arm, vitamin D alone was associated with an HR for incident autoimmune disease of 0.68 (P = .02), omega-3 alone was associated with a nonsignificant HR of 0.74, and the combination was associated with an HR of 0.69 (P = .03).

Dr. Costenbader and colleagues acknowledged that the study was limited by the lack of a high-risk or nutritionally-deficient population, where the effects of supplementation might be larger; the restriction of the sample to older adults; and to the difficulty of confirming incident autoimmune thyroid disease from patient reports.

Cheryl Koehn, an arthritis patient advocate from Vancouver, Canada, who was not involved in the study, commented in the “chat” section of the presentation that her rheumatologist “has recommended vitamin D for years now. Says basically everyone north of Boston is vitamin D deficient. I take 1,000 IU per day. Been taking it for years.” Ms. Koehn is the founder and president of Arthritis Consumer Experts, a website that provides education to those with arthritis.

“Agreed. I tell every patient to take vitamin D supplement,” commented Fatma Dedeoglu, MD, a rheumatologist at Boston Children’s Hospital.



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

For those of us who cannot sit in the sun and fish all day, the next best thing for preventing autoimmune diseases may be supplementation with vitamin D and fish oil-derived omega-3 fatty acids, results of a large prospective randomized trial suggest.

Ziga Plahutar

Among nearly 26,000 adults enrolled in a randomized trial designed primarily to study the effects of vitamin D and omega-3 supplementation on incident cancer and cardiovascular disease, 5 years of vitamin D supplementation was associated with a 22% reduction in risk for confirmed autoimmune diseases, and 5 years of omega-3 fatty acid supplementation was associated with an 18% reduction in confirmed and probable incident autoimmune diseases, reported Karen H. Costenbader, MD, MPH, of Brigham & Women’s Hospital in Boston.

“The clinical importance of these results is very high, given that these are nontoxic, well-tolerated supplements, and that there are no other known effective therapies to reduce the incidence of autoimmune diseases,” she said during the virtual annual meeting of the American College of Rheumatology.

“People do have to take the supplements a long time to start to see the reduction in risk, especially for vitamin D, but they make biological sense, and autoimmune diseases develop slowly over time, so taking it today isn’t going to reduce risk of developing something tomorrow,” Dr. Costenbader said in an interview.

“These supplements have other health benefits. Obviously, fish oil is anti-inflammatory, and vitamin D is good for osteoporosis prevention, especially in our patients who take glucocorticoids. People who are otherwise healthy and have a family history of autoimmune disease might also consider starting to take these supplements,” she said.

After watching her presentation, session co-moderator Gregg Silverman, MD, from the NYU Langone School of Medicine in New York, who was not involved in the study, commented “I’m going to [nutrition store] GNC to get some vitamins.”

When asked for comment, the other session moderator, Tracy Frech, MD, of Vanderbilt University, Nashville, said, “I think Dr. Costenbader’s work is very important and her presentation excellent. My current practice is replacement of vitamin D in all autoimmune disease patients with low levels and per bone health guidelines. Additionally, I discuss omega-3 supplementation with Sjögren’s [syndrome] patients as a consideration.”

Evidence base

Dr. Costenbader noted that in a 2013 observational study from France, vitamin D derived through ultraviolet (UV) light exposure was associated with a lower risk for incident Crohn’s disease but not ulcerative colitis, and in two analyses of data in 2014 from the Nurses’ Health Study, both high plasma levels of 25-OH vitamin D and geographic residence in areas of high UV exposure were associated with a decreased incidence of rheumatoid arthritis (RA).

Other observational studies have supported omega-3 fatty acids for their anti-inflammatory properties, including a 2005 Danish prospective cohort study showing a lower risk for RA in participants who reported higher levels of fatty fish intake. In a separate study conducted in 2017, healthy volunteers with higher omega-3 fatty acid/total lipid proportions in red blood cell membranes had a lower prevalence of anti-cyclic citrullinated peptide (anti-CCP) antibodies and rheumatoid factor and a lower incidence of progression to inflammatory arthritis, she said.

Ancillary study

Despite the evidence, however, there have been no prospective randomized trials to test the effects of either vitamin D or omega-3 fatty acid supplementation on the incidence of autoimmune disease over time.

To rectify this, Dr. Costenbader and colleagues piggybacked an ancillary study onto the Vitamin D and Omega-3 Trial (VITAL), which had primary outcomes of cancer and cardiovascular disease incidence.

A total of 25,871 participants were enrolled, including 12,786 men aged 50 and older, and 13,085 women aged 55 and older.

The study had a 2 x 2 factorial design, with patients randomly assigned to vitamin D 2,000 IU/day or placebo, and then further randomized to either 1 g/day omega-3 fatty acids or placebo in both the vitamin D and placebo primary randomization arms.

At baseline 16,956 participants were assayed for 25-OH vitamin D and plasma omega 3 index, the ratio of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) to total fatty acids. Participants self-reported baseline and all incident autoimmune diseases annually, with the reports confirmed by medical record review and disease criteria whenever possible.

Results

At 5 years of follow-up, confirmed incident autoimmune diseases had occurred in 123 patients in the active vitamin D group, compared with 155 in the placebo vitamin D group, translating into a hazard ratio (HR) for vitamin D of 0.78 (P = .045).

In the active omega-3 arm, 130 participants developed an autoimmune disease, compared with 148 in the placebo omega-3 arm, which translated into a nonsignificant HR of 0.85.

There was no statistical interaction between the two supplements. The investigators did observe an interaction between vitamin D and body mass index, with the effect stronger among participants with low BMI (P = .02). There also was an interaction between omega-3 fatty acids with a family history of autoimmune disease (P = .03).

In multivariate analysis adjusted for age, sex, race, and other supplement arm, vitamin D alone was associated with an HR for incident autoimmune disease of 0.68 (P = .02), omega-3 alone was associated with a nonsignificant HR of 0.74, and the combination was associated with an HR of 0.69 (P = .03).

Dr. Costenbader and colleagues acknowledged that the study was limited by the lack of a high-risk or nutritionally-deficient population, where the effects of supplementation might be larger; the restriction of the sample to older adults; and to the difficulty of confirming incident autoimmune thyroid disease from patient reports.

Cheryl Koehn, an arthritis patient advocate from Vancouver, Canada, who was not involved in the study, commented in the “chat” section of the presentation that her rheumatologist “has recommended vitamin D for years now. Says basically everyone north of Boston is vitamin D deficient. I take 1,000 IU per day. Been taking it for years.” Ms. Koehn is the founder and president of Arthritis Consumer Experts, a website that provides education to those with arthritis.

“Agreed. I tell every patient to take vitamin D supplement,” commented Fatma Dedeoglu, MD, a rheumatologist at Boston Children’s Hospital.



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

For those of us who cannot sit in the sun and fish all day, the next best thing for preventing autoimmune diseases may be supplementation with vitamin D and fish oil-derived omega-3 fatty acids, results of a large prospective randomized trial suggest.

Ziga Plahutar

Among nearly 26,000 adults enrolled in a randomized trial designed primarily to study the effects of vitamin D and omega-3 supplementation on incident cancer and cardiovascular disease, 5 years of vitamin D supplementation was associated with a 22% reduction in risk for confirmed autoimmune diseases, and 5 years of omega-3 fatty acid supplementation was associated with an 18% reduction in confirmed and probable incident autoimmune diseases, reported Karen H. Costenbader, MD, MPH, of Brigham & Women’s Hospital in Boston.

“The clinical importance of these results is very high, given that these are nontoxic, well-tolerated supplements, and that there are no other known effective therapies to reduce the incidence of autoimmune diseases,” she said during the virtual annual meeting of the American College of Rheumatology.

“People do have to take the supplements a long time to start to see the reduction in risk, especially for vitamin D, but they make biological sense, and autoimmune diseases develop slowly over time, so taking it today isn’t going to reduce risk of developing something tomorrow,” Dr. Costenbader said in an interview.

“These supplements have other health benefits. Obviously, fish oil is anti-inflammatory, and vitamin D is good for osteoporosis prevention, especially in our patients who take glucocorticoids. People who are otherwise healthy and have a family history of autoimmune disease might also consider starting to take these supplements,” she said.

After watching her presentation, session co-moderator Gregg Silverman, MD, from the NYU Langone School of Medicine in New York, who was not involved in the study, commented “I’m going to [nutrition store] GNC to get some vitamins.”

When asked for comment, the other session moderator, Tracy Frech, MD, of Vanderbilt University, Nashville, said, “I think Dr. Costenbader’s work is very important and her presentation excellent. My current practice is replacement of vitamin D in all autoimmune disease patients with low levels and per bone health guidelines. Additionally, I discuss omega-3 supplementation with Sjögren’s [syndrome] patients as a consideration.”

Evidence base

Dr. Costenbader noted that in a 2013 observational study from France, vitamin D derived through ultraviolet (UV) light exposure was associated with a lower risk for incident Crohn’s disease but not ulcerative colitis, and in two analyses of data in 2014 from the Nurses’ Health Study, both high plasma levels of 25-OH vitamin D and geographic residence in areas of high UV exposure were associated with a decreased incidence of rheumatoid arthritis (RA).

Other observational studies have supported omega-3 fatty acids for their anti-inflammatory properties, including a 2005 Danish prospective cohort study showing a lower risk for RA in participants who reported higher levels of fatty fish intake. In a separate study conducted in 2017, healthy volunteers with higher omega-3 fatty acid/total lipid proportions in red blood cell membranes had a lower prevalence of anti-cyclic citrullinated peptide (anti-CCP) antibodies and rheumatoid factor and a lower incidence of progression to inflammatory arthritis, she said.

Ancillary study

Despite the evidence, however, there have been no prospective randomized trials to test the effects of either vitamin D or omega-3 fatty acid supplementation on the incidence of autoimmune disease over time.

To rectify this, Dr. Costenbader and colleagues piggybacked an ancillary study onto the Vitamin D and Omega-3 Trial (VITAL), which had primary outcomes of cancer and cardiovascular disease incidence.

A total of 25,871 participants were enrolled, including 12,786 men aged 50 and older, and 13,085 women aged 55 and older.

The study had a 2 x 2 factorial design, with patients randomly assigned to vitamin D 2,000 IU/day or placebo, and then further randomized to either 1 g/day omega-3 fatty acids or placebo in both the vitamin D and placebo primary randomization arms.

At baseline 16,956 participants were assayed for 25-OH vitamin D and plasma omega 3 index, the ratio of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) to total fatty acids. Participants self-reported baseline and all incident autoimmune diseases annually, with the reports confirmed by medical record review and disease criteria whenever possible.

Results

At 5 years of follow-up, confirmed incident autoimmune diseases had occurred in 123 patients in the active vitamin D group, compared with 155 in the placebo vitamin D group, translating into a hazard ratio (HR) for vitamin D of 0.78 (P = .045).

In the active omega-3 arm, 130 participants developed an autoimmune disease, compared with 148 in the placebo omega-3 arm, which translated into a nonsignificant HR of 0.85.

There was no statistical interaction between the two supplements. The investigators did observe an interaction between vitamin D and body mass index, with the effect stronger among participants with low BMI (P = .02). There also was an interaction between omega-3 fatty acids with a family history of autoimmune disease (P = .03).

In multivariate analysis adjusted for age, sex, race, and other supplement arm, vitamin D alone was associated with an HR for incident autoimmune disease of 0.68 (P = .02), omega-3 alone was associated with a nonsignificant HR of 0.74, and the combination was associated with an HR of 0.69 (P = .03).

Dr. Costenbader and colleagues acknowledged that the study was limited by the lack of a high-risk or nutritionally-deficient population, where the effects of supplementation might be larger; the restriction of the sample to older adults; and to the difficulty of confirming incident autoimmune thyroid disease from patient reports.

Cheryl Koehn, an arthritis patient advocate from Vancouver, Canada, who was not involved in the study, commented in the “chat” section of the presentation that her rheumatologist “has recommended vitamin D for years now. Says basically everyone north of Boston is vitamin D deficient. I take 1,000 IU per day. Been taking it for years.” Ms. Koehn is the founder and president of Arthritis Consumer Experts, a website that provides education to those with arthritis.

“Agreed. I tell every patient to take vitamin D supplement,” commented Fatma Dedeoglu, MD, a rheumatologist at Boston Children’s Hospital.



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

Identifying the molecular features of metastatic breast cancer (MBC) offers a real-time window into a patient’s treatment options as well as the potential to follow the disease as it evolves over time.
 

Tissue biopsy remains the gold standard for characterizing tumor biology and guiding therapeutic decisions, but liquid biopsies — blood analyses that allow oncologists to detect circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) in the blood — are increasingly demonstrating their value. Last year, the U.S. Food and Drug Administration (FDA) approved two liquid biopsy tests, Guardant360 CDx and FoundationOne Liquid CDx, that can identify more than 300 cancer-related genes in the blood. In 2019, the FDA also approved the first companion diagnostic test, therascreen, to pinpoint PIK3CA gene mutations in patients’ ctDNA and determine whether patients should receive the PI3K inhibitor alpelisib along with fulvestrant.

Here’s an overview of how liquid biopsy is being used in monitoring MBC progression and treatment — and what some oncologists think of it.

What we do and don’t know

“Identifying a patient’s targetable mutations, most notably PIK3CA mutations, is currently the main use of liquid biopsy,” said Pedram Razavi, MD, PhD, a medical oncologist who leads the liquid biopsy program for breast cancer at Memorial Sloan Kettering (MSK) Cancer Center in New York City. “Patients who come to MSK are offered a tumor and liquid biopsy at the time of metastatic diagnosis as part of the standard of care.”

Liquid and tissue biopsy analyses can provide a more complete picture of a patient’s condition. Whereas tissue biopsy allows oncologists to target a more saturated sample of the cancer ecosystem and a wider array of biomarkers, liquid biopsy offers important advantages as well, including a less invasive way to sequence a sample, monitor patients’ treatment response, or track tumor evolution. Liquid biopsy also provides a bigger picture view of tumor heterogeneity by pooling information from many tumor locations as opposed to one.

But, cautioned Yuan Yuan, MD, PhD, liquid biopsy technology is not always sensitive enough to detect CTCs, ctDNA, or all relevant mutations. “When you collect a small tube of blood, you’re essentially trying to catch a small fish in a big sea and wading through a lot of background noise,” said Dr. Yuan, medical oncologist at City of Hope, a comprehensive cancer center in Los Angeles County. “The results may be hard to interpret or come back inconclusive.”

And although emerging data suggest that liquid biopsy provides important insights about tumor dynamics — including mapping disease progression, predicting survival, and even detecting signs of cancer recurrence before metastasis develops — the tool has limited utility in clinical practice outside of identifying sensitivity to various therapies or drugs.

“Right now, a lot of research is being done to understand how to use CTC and ctDNA in particular as a means of surveillance in breast cancer, but we’re still in the beginning stages of applying that outside of clinical trials,” said Joseph A. Sparano, MD, deputy director of the Tisch Cancer Institute and chief of the division of hematology and medical oncology, Icahn School of Medicine at Mount Sinai, New York City.

Personalizing treatment

The companion diagnostic test therascreen marked the beginning stages of using liquid biopsy to match treatments to genetic abnormalities in MBC. The SOLAR-1 phase 3 trial, which led to the approval of alpelisib and therascreen, found that the PI3K inhibitor plus fulvestrant almost doubled progression-free survival (PFS) (11 months vs 5.7 months in placebo-fulvestrant group) in patients with PIK3CA-mutated, HR-positive, HER2-negative advanced breast cancer.

More recent studies have shown that liquid biopsy tests can also identify ESR1 mutations and predict responses to inhibitors that target AKT1 and HER2. Investigators presenting at the 2021 American Society of Clinical Oncology meeting reported that next-generation sequencing of ctDNA in patients with HR-positive MBC, HER-positive MBC, or triple-negative breast cancer detected ESR1 mutations in 14% of patients (71 of 501). Moreover, ESR1 mutations were found only in HR-positive patients who had already received endocrine therapy. (The study also examined PIK3CA mutations, which occurred in about one third of patients). A more in-depth look revealed that ESR1 mutations were strongly associated with liver and bone metastases and that mutations along specific codons negatively affected overall survival (OS) and PFS: codons 537 and 538 for OS and codons 380 and 536 for PFS.

According to Debasish Tripathy, MD, professor and chairman of the department of breast medical oncology at the University of Texas MD Anderson Cancer Center in Houston, in addition to tumor sequencing, “liquid biopsy has become a great research tool to track patients in real time and predict, for instance, who will respond to a treatment and identify emerging resistance.”

In terms of predicting responses to treatment, the plasmaMATCH trial assessed ctDNA in 1,034 patients with advanced breast cancer for mutations in ESR1, HER2, and AKT1 using digital droplet polymerase chain reaction (PCR) and Guardant360. Results showed that 357 (34.5%) of these patients had potentially targetable aberrations, including 222 patients with ESR1 mutations, 36 patients with HER2 mutations, and 30 patients with AKT1 mutations.

Agreement between digital droplet PCR and Guardant360 testing was 96%-99%, and liquid biopsy showed 93% sensitivity compared with tumor samples. The investigators also used liquid biopsy findings to match patients’ mutations to targeted treatments: fulvestrant for those with ESR1 mutations, neratinib for HER2 (ERBB2) mutations, and the selective AKT inhibitor capivasertib for estrogen receptor–positive tumors with AKT1 mutations.

Overall, the investigators concluded that ctDNA testing offers “accurate tumor genotyping” in line with tissue-based testing and is ready for routine clinical practice to identify common as well as rare genetic alterations, such as HER2 and AKT1 mutations, that affect only about 5% of patients with advanced disease.

Predicting survival and recurrence

A particularly promising area for liquid biopsy is its usefulness in helping to predict survival outcomes and monitor patients for early signs of recurrence before metastasis occurs. But the data to support this are still in their infancy.

A highly cited study, published over 15 years ago in the New England Journal of Medicine, found that patients with MBC who had five or more CTCs per 7.5 mL of whole blood before receiving first-line therapy exhibited significantly shorter median PFS (2.7 vs 7.0 months) and OS (10 vs > 18 months) compared with patients with fewer than five CTCs. Subsequent analyses performed more than a decade later, including a meta-analysis published last year, helped validate these early findings that levels of CTCs detected in the blood independently and strongly predicted PFS and OS in patients with MBC.

In addition, ctDNA can provide important information about patients’ survival odds. In a retrospective study published last year, investigators tracked changes in ctDNA in 291 plasma samples from 84 patients with locally advanced breast cancer who participated in the I-SPY trial. Patients who remained ctDNA-positive after 3 weeks of neoadjuvant chemotherapy were significantly more likely to have residual disease after completing their treatment compared with patients who cleared ctDNA at that early stage (83% for those with nonpathologic complete response vs 52%). Notably, the presence of ctDNA between therapy initiation and completion was associated with a significantly greater risk for metastatic recurrence, whereas clearance of ctDNA after neoadjuvant therapy was linked to improved survival.

“The study is important because it highlights how tracking circulating ctDNA status in neoadjuvant-treated breast cancer can expose a patient’s risk for distant metastasis,” said Dr. Yuan. But, she added, “I think the biggest attraction of liquid biopsy will be the ability to detect molecular disease even before imaging can, and identify who has a high risk for recurrence.”

Dr. Razavi agreed that the potential to prevent metastasis by finding minimal residual disease (MRD) is the most exciting area of liquid biopsy research. “If we can find tumor DNA early before tumors have a chance to establish themselves, we could potentially change the trajectory of the disease for patients,” he said.

Several studies suggest that monitoring patients’ ctDNA levels after neoadjuvant treatment and surgery may help predict their risk for relapse and progression to metastatic disease. A 2015 analysis, which followed 20 patients with breast cancer after surgery, found that ctDNA monitoring accurately differentiated those who ultimately developed metastatic disease from those who didn’t (sensitivity, 93%; specificity, 100%) and detected metastatic disease 11 months earlier, on average, than imaging did. Another 2015 study found that the presence of ctDNA in plasma after neoadjuvant chemotherapy and surgery predicted metastatic relapse a median of almost 8 months before clinical detection. Other recent data show the power of ultrasensitive blood tests to detect MRD and potentially find metastatic disease early.

Although an increasing number of studies show that ctDNA and CTCs are prognostic for breast cancer recurrence, a major question remains: For patients with ctDNA or CTCs but no overt disease after imaging, will initiating therapy prevent or delay the development of metastatic disease?

“We still have to do those clinical trials to determine whether detecting MRD and treating patients early actually positively affects their survival and quality of life,” Dr. Razavi said.

Identifying the molecular features of metastatic breast cancer (MBC) offers a real-time window into a patient’s treatment options as well as the potential to follow the disease as it evolves over time.
 

Tissue biopsy remains the gold standard for characterizing tumor biology and guiding therapeutic decisions, but liquid biopsies — blood analyses that allow oncologists to detect circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) in the blood — are increasingly demonstrating their value. Last year, the U.S. Food and Drug Administration (FDA) approved two liquid biopsy tests, Guardant360 CDx and FoundationOne Liquid CDx, that can identify more than 300 cancer-related genes in the blood. In 2019, the FDA also approved the first companion diagnostic test, therascreen, to pinpoint PIK3CA gene mutations in patients’ ctDNA and determine whether patients should receive the PI3K inhibitor alpelisib along with fulvestrant.

Here’s an overview of how liquid biopsy is being used in monitoring MBC progression and treatment — and what some oncologists think of it.

What we do and don’t know

“Identifying a patient’s targetable mutations, most notably PIK3CA mutations, is currently the main use of liquid biopsy,” said Pedram Razavi, MD, PhD, a medical oncologist who leads the liquid biopsy program for breast cancer at Memorial Sloan Kettering (MSK) Cancer Center in New York City. “Patients who come to MSK are offered a tumor and liquid biopsy at the time of metastatic diagnosis as part of the standard of care.”

Liquid and tissue biopsy analyses can provide a more complete picture of a patient’s condition. Whereas tissue biopsy allows oncologists to target a more saturated sample of the cancer ecosystem and a wider array of biomarkers, liquid biopsy offers important advantages as well, including a less invasive way to sequence a sample, monitor patients’ treatment response, or track tumor evolution. Liquid biopsy also provides a bigger picture view of tumor heterogeneity by pooling information from many tumor locations as opposed to one.

But, cautioned Yuan Yuan, MD, PhD, liquid biopsy technology is not always sensitive enough to detect CTCs, ctDNA, or all relevant mutations. “When you collect a small tube of blood, you’re essentially trying to catch a small fish in a big sea and wading through a lot of background noise,” said Dr. Yuan, medical oncologist at City of Hope, a comprehensive cancer center in Los Angeles County. “The results may be hard to interpret or come back inconclusive.”

And although emerging data suggest that liquid biopsy provides important insights about tumor dynamics — including mapping disease progression, predicting survival, and even detecting signs of cancer recurrence before metastasis develops — the tool has limited utility in clinical practice outside of identifying sensitivity to various therapies or drugs.

“Right now, a lot of research is being done to understand how to use CTC and ctDNA in particular as a means of surveillance in breast cancer, but we’re still in the beginning stages of applying that outside of clinical trials,” said Joseph A. Sparano, MD, deputy director of the Tisch Cancer Institute and chief of the division of hematology and medical oncology, Icahn School of Medicine at Mount Sinai, New York City.

Personalizing treatment

The companion diagnostic test therascreen marked the beginning stages of using liquid biopsy to match treatments to genetic abnormalities in MBC. The SOLAR-1 phase 3 trial, which led to the approval of alpelisib and therascreen, found that the PI3K inhibitor plus fulvestrant almost doubled progression-free survival (PFS) (11 months vs 5.7 months in placebo-fulvestrant group) in patients with PIK3CA-mutated, HR-positive, HER2-negative advanced breast cancer.

More recent studies have shown that liquid biopsy tests can also identify ESR1 mutations and predict responses to inhibitors that target AKT1 and HER2. Investigators presenting at the 2021 American Society of Clinical Oncology meeting reported that next-generation sequencing of ctDNA in patients with HR-positive MBC, HER-positive MBC, or triple-negative breast cancer detected ESR1 mutations in 14% of patients (71 of 501). Moreover, ESR1 mutations were found only in HR-positive patients who had already received endocrine therapy. (The study also examined PIK3CA mutations, which occurred in about one third of patients). A more in-depth look revealed that ESR1 mutations were strongly associated with liver and bone metastases and that mutations along specific codons negatively affected overall survival (OS) and PFS: codons 537 and 538 for OS and codons 380 and 536 for PFS.

According to Debasish Tripathy, MD, professor and chairman of the department of breast medical oncology at the University of Texas MD Anderson Cancer Center in Houston, in addition to tumor sequencing, “liquid biopsy has become a great research tool to track patients in real time and predict, for instance, who will respond to a treatment and identify emerging resistance.”

In terms of predicting responses to treatment, the plasmaMATCH trial assessed ctDNA in 1,034 patients with advanced breast cancer for mutations in ESR1, HER2, and AKT1 using digital droplet polymerase chain reaction (PCR) and Guardant360. Results showed that 357 (34.5%) of these patients had potentially targetable aberrations, including 222 patients with ESR1 mutations, 36 patients with HER2 mutations, and 30 patients with AKT1 mutations.

Agreement between digital droplet PCR and Guardant360 testing was 96%-99%, and liquid biopsy showed 93% sensitivity compared with tumor samples. The investigators also used liquid biopsy findings to match patients’ mutations to targeted treatments: fulvestrant for those with ESR1 mutations, neratinib for HER2 (ERBB2) mutations, and the selective AKT inhibitor capivasertib for estrogen receptor–positive tumors with AKT1 mutations.

Overall, the investigators concluded that ctDNA testing offers “accurate tumor genotyping” in line with tissue-based testing and is ready for routine clinical practice to identify common as well as rare genetic alterations, such as HER2 and AKT1 mutations, that affect only about 5% of patients with advanced disease.

Predicting survival and recurrence

A particularly promising area for liquid biopsy is its usefulness in helping to predict survival outcomes and monitor patients for early signs of recurrence before metastasis occurs. But the data to support this are still in their infancy.

A highly cited study, published over 15 years ago in the New England Journal of Medicine, found that patients with MBC who had five or more CTCs per 7.5 mL of whole blood before receiving first-line therapy exhibited significantly shorter median PFS (2.7 vs 7.0 months) and OS (10 vs > 18 months) compared with patients with fewer than five CTCs. Subsequent analyses performed more than a decade later, including a meta-analysis published last year, helped validate these early findings that levels of CTCs detected in the blood independently and strongly predicted PFS and OS in patients with MBC.

In addition, ctDNA can provide important information about patients’ survival odds. In a retrospective study published last year, investigators tracked changes in ctDNA in 291 plasma samples from 84 patients with locally advanced breast cancer who participated in the I-SPY trial. Patients who remained ctDNA-positive after 3 weeks of neoadjuvant chemotherapy were significantly more likely to have residual disease after completing their treatment compared with patients who cleared ctDNA at that early stage (83% for those with nonpathologic complete response vs 52%). Notably, the presence of ctDNA between therapy initiation and completion was associated with a significantly greater risk for metastatic recurrence, whereas clearance of ctDNA after neoadjuvant therapy was linked to improved survival.

“The study is important because it highlights how tracking circulating ctDNA status in neoadjuvant-treated breast cancer can expose a patient’s risk for distant metastasis,” said Dr. Yuan. But, she added, “I think the biggest attraction of liquid biopsy will be the ability to detect molecular disease even before imaging can, and identify who has a high risk for recurrence.”

Dr. Razavi agreed that the potential to prevent metastasis by finding minimal residual disease (MRD) is the most exciting area of liquid biopsy research. “If we can find tumor DNA early before tumors have a chance to establish themselves, we could potentially change the trajectory of the disease for patients,” he said.

Several studies suggest that monitoring patients’ ctDNA levels after neoadjuvant treatment and surgery may help predict their risk for relapse and progression to metastatic disease. A 2015 analysis, which followed 20 patients with breast cancer after surgery, found that ctDNA monitoring accurately differentiated those who ultimately developed metastatic disease from those who didn’t (sensitivity, 93%; specificity, 100%) and detected metastatic disease 11 months earlier, on average, than imaging did. Another 2015 study found that the presence of ctDNA in plasma after neoadjuvant chemotherapy and surgery predicted metastatic relapse a median of almost 8 months before clinical detection. Other recent data show the power of ultrasensitive blood tests to detect MRD and potentially find metastatic disease early.

Although an increasing number of studies show that ctDNA and CTCs are prognostic for breast cancer recurrence, a major question remains: For patients with ctDNA or CTCs but no overt disease after imaging, will initiating therapy prevent or delay the development of metastatic disease?

“We still have to do those clinical trials to determine whether detecting MRD and treating patients early actually positively affects their survival and quality of life,” Dr. Razavi said.

Identifying the molecular features of metastatic breast cancer (MBC) offers a real-time window into a patient’s treatment options as well as the potential to follow the disease as it evolves over time.
 

Tissue biopsy remains the gold standard for characterizing tumor biology and guiding therapeutic decisions, but liquid biopsies — blood analyses that allow oncologists to detect circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) in the blood — are increasingly demonstrating their value. Last year, the U.S. Food and Drug Administration (FDA) approved two liquid biopsy tests, Guardant360 CDx and FoundationOne Liquid CDx, that can identify more than 300 cancer-related genes in the blood. In 2019, the FDA also approved the first companion diagnostic test, therascreen, to pinpoint PIK3CA gene mutations in patients’ ctDNA and determine whether patients should receive the PI3K inhibitor alpelisib along with fulvestrant.

Here’s an overview of how liquid biopsy is being used in monitoring MBC progression and treatment — and what some oncologists think of it.

What we do and don’t know

“Identifying a patient’s targetable mutations, most notably PIK3CA mutations, is currently the main use of liquid biopsy,” said Pedram Razavi, MD, PhD, a medical oncologist who leads the liquid biopsy program for breast cancer at Memorial Sloan Kettering (MSK) Cancer Center in New York City. “Patients who come to MSK are offered a tumor and liquid biopsy at the time of metastatic diagnosis as part of the standard of care.”

Liquid and tissue biopsy analyses can provide a more complete picture of a patient’s condition. Whereas tissue biopsy allows oncologists to target a more saturated sample of the cancer ecosystem and a wider array of biomarkers, liquid biopsy offers important advantages as well, including a less invasive way to sequence a sample, monitor patients’ treatment response, or track tumor evolution. Liquid biopsy also provides a bigger picture view of tumor heterogeneity by pooling information from many tumor locations as opposed to one.

But, cautioned Yuan Yuan, MD, PhD, liquid biopsy technology is not always sensitive enough to detect CTCs, ctDNA, or all relevant mutations. “When you collect a small tube of blood, you’re essentially trying to catch a small fish in a big sea and wading through a lot of background noise,” said Dr. Yuan, medical oncologist at City of Hope, a comprehensive cancer center in Los Angeles County. “The results may be hard to interpret or come back inconclusive.”

And although emerging data suggest that liquid biopsy provides important insights about tumor dynamics — including mapping disease progression, predicting survival, and even detecting signs of cancer recurrence before metastasis develops — the tool has limited utility in clinical practice outside of identifying sensitivity to various therapies or drugs.

“Right now, a lot of research is being done to understand how to use CTC and ctDNA in particular as a means of surveillance in breast cancer, but we’re still in the beginning stages of applying that outside of clinical trials,” said Joseph A. Sparano, MD, deputy director of the Tisch Cancer Institute and chief of the division of hematology and medical oncology, Icahn School of Medicine at Mount Sinai, New York City.

Personalizing treatment

The companion diagnostic test therascreen marked the beginning stages of using liquid biopsy to match treatments to genetic abnormalities in MBC. The SOLAR-1 phase 3 trial, which led to the approval of alpelisib and therascreen, found that the PI3K inhibitor plus fulvestrant almost doubled progression-free survival (PFS) (11 months vs 5.7 months in placebo-fulvestrant group) in patients with PIK3CA-mutated, HR-positive, HER2-negative advanced breast cancer.

More recent studies have shown that liquid biopsy tests can also identify ESR1 mutations and predict responses to inhibitors that target AKT1 and HER2. Investigators presenting at the 2021 American Society of Clinical Oncology meeting reported that next-generation sequencing of ctDNA in patients with HR-positive MBC, HER-positive MBC, or triple-negative breast cancer detected ESR1 mutations in 14% of patients (71 of 501). Moreover, ESR1 mutations were found only in HR-positive patients who had already received endocrine therapy. (The study also examined PIK3CA mutations, which occurred in about one third of patients). A more in-depth look revealed that ESR1 mutations were strongly associated with liver and bone metastases and that mutations along specific codons negatively affected overall survival (OS) and PFS: codons 537 and 538 for OS and codons 380 and 536 for PFS.

According to Debasish Tripathy, MD, professor and chairman of the department of breast medical oncology at the University of Texas MD Anderson Cancer Center in Houston, in addition to tumor sequencing, “liquid biopsy has become a great research tool to track patients in real time and predict, for instance, who will respond to a treatment and identify emerging resistance.”

In terms of predicting responses to treatment, the plasmaMATCH trial assessed ctDNA in 1,034 patients with advanced breast cancer for mutations in ESR1, HER2, and AKT1 using digital droplet polymerase chain reaction (PCR) and Guardant360. Results showed that 357 (34.5%) of these patients had potentially targetable aberrations, including 222 patients with ESR1 mutations, 36 patients with HER2 mutations, and 30 patients with AKT1 mutations.

Agreement between digital droplet PCR and Guardant360 testing was 96%-99%, and liquid biopsy showed 93% sensitivity compared with tumor samples. The investigators also used liquid biopsy findings to match patients’ mutations to targeted treatments: fulvestrant for those with ESR1 mutations, neratinib for HER2 (ERBB2) mutations, and the selective AKT inhibitor capivasertib for estrogen receptor–positive tumors with AKT1 mutations.

Overall, the investigators concluded that ctDNA testing offers “accurate tumor genotyping” in line with tissue-based testing and is ready for routine clinical practice to identify common as well as rare genetic alterations, such as HER2 and AKT1 mutations, that affect only about 5% of patients with advanced disease.

Predicting survival and recurrence

A particularly promising area for liquid biopsy is its usefulness in helping to predict survival outcomes and monitor patients for early signs of recurrence before metastasis occurs. But the data to support this are still in their infancy.

A highly cited study, published over 15 years ago in the New England Journal of Medicine, found that patients with MBC who had five or more CTCs per 7.5 mL of whole blood before receiving first-line therapy exhibited significantly shorter median PFS (2.7 vs 7.0 months) and OS (10 vs > 18 months) compared with patients with fewer than five CTCs. Subsequent analyses performed more than a decade later, including a meta-analysis published last year, helped validate these early findings that levels of CTCs detected in the blood independently and strongly predicted PFS and OS in patients with MBC.

In addition, ctDNA can provide important information about patients’ survival odds. In a retrospective study published last year, investigators tracked changes in ctDNA in 291 plasma samples from 84 patients with locally advanced breast cancer who participated in the I-SPY trial. Patients who remained ctDNA-positive after 3 weeks of neoadjuvant chemotherapy were significantly more likely to have residual disease after completing their treatment compared with patients who cleared ctDNA at that early stage (83% for those with nonpathologic complete response vs 52%). Notably, the presence of ctDNA between therapy initiation and completion was associated with a significantly greater risk for metastatic recurrence, whereas clearance of ctDNA after neoadjuvant therapy was linked to improved survival.

“The study is important because it highlights how tracking circulating ctDNA status in neoadjuvant-treated breast cancer can expose a patient’s risk for distant metastasis,” said Dr. Yuan. But, she added, “I think the biggest attraction of liquid biopsy will be the ability to detect molecular disease even before imaging can, and identify who has a high risk for recurrence.”

Dr. Razavi agreed that the potential to prevent metastasis by finding minimal residual disease (MRD) is the most exciting area of liquid biopsy research. “If we can find tumor DNA early before tumors have a chance to establish themselves, we could potentially change the trajectory of the disease for patients,” he said.

Several studies suggest that monitoring patients’ ctDNA levels after neoadjuvant treatment and surgery may help predict their risk for relapse and progression to metastatic disease. A 2015 analysis, which followed 20 patients with breast cancer after surgery, found that ctDNA monitoring accurately differentiated those who ultimately developed metastatic disease from those who didn’t (sensitivity, 93%; specificity, 100%) and detected metastatic disease 11 months earlier, on average, than imaging did. Another 2015 study found that the presence of ctDNA in plasma after neoadjuvant chemotherapy and surgery predicted metastatic relapse a median of almost 8 months before clinical detection. Other recent data show the power of ultrasensitive blood tests to detect MRD and potentially find metastatic disease early.

Although an increasing number of studies show that ctDNA and CTCs are prognostic for breast cancer recurrence, a major question remains: For patients with ctDNA or CTCs but no overt disease after imaging, will initiating therapy prevent or delay the development of metastatic disease?

“We still have to do those clinical trials to determine whether detecting MRD and treating patients early actually positively affects their survival and quality of life,” Dr. Razavi said.

Artificial Intelligence (AI) was first described in 1956 and refers to machines having the ability to learn as they receive and process information, resulting in the ability to “think” like humans.¹ AI’s impact in medicine is increasing; currently, at least 29 AI medical devices and algorithms are approved by the US Food and Drug Administration (FDA) in a variety of areas, including radiograph interpretation, managing glucose levels in patients with diabetes mellitus, analyzing electrocardiograms (ECGs), and diagnosing sleep disorders among others.² Significantly, in 2020, the Centers for Medicare and Medicaid Services (CMS) announced the first reimbursement to hospitals for an AI platform, a model for early detection of strokes.³ AI is rapidly becoming an integral part of health care, and its role will only increase in the future (Table).

As knowledge in medicine is expanding exponentially, AI has great potential to assist with handling complex patient care data. The concept of exponential growth is not a natural one. As Bini described, with exponential growth the volume of knowledge amassed over the past 10 years will now occur in perhaps only 1 year.¹ Likewise, equivalent advances over the past year may take just a few months. This phenomenon is partly due to the law of accelerating returns, which states that advances feed on themselves, continually increasing the rate of further advances.⁴ The volume of medical data doubles every 2 to 5 years.⁵ Fortunately, the field of AI is growing exponentially as well and can help health care practitioners (HCPs) keep pace, allowing the continued delivery of effective health care.

In this report, we review common terminology, principles, and general applications of AI, followed by current and potential applications of AI for selected medical specialties. Finally, we discuss AI’s future in health care, along with potential risks and pitfalls.

AI Overview

AI refers to machine programs that can “learn” or think based on past experiences. This functionality contrasts with simple rules-based programming available to health care for years. An example of rules-based programming is the warfarindosing.org website developed by Barnes-Jewish Hospital at Washington University Medical Center, which guides initial warfarin dosing.^6,7 The prescriber inputs detailed patient information, including age, sex, height, weight, tobacco history, medications, laboratory results, and genotype if available. The application then calculates recommended warfarin dosing regimens to avoid over- or underanticoagulation. While the dosing algorithm may be complex, it depends entirely on preprogrammed rules. The program does not learn to reach its conclusions and recommendations from patient data.

In contrast, one of the most common subsets of AI is machine learning (ML). ML describes a program that “learns from experience and improves its performance as it learns.”¹ With ML, the computer is initially provided with a training data set—data with known outcomes or labels. Because the initial data are input from known samples, this type of AI is known as supervised learning.^8-10 As an example, we recently reported using ML to diagnose various types of cancer from pathology slides.¹¹ In one experiment, we captured images of colon adenocarcinoma and normal colon (these 2 groups represent the training data set). Unlike traditional programming, we did not define characteristics that would differentiate colon cancer from normal; rather, the machine learned these characteristics independently by assessing the labeled images provided. A second data set (the validation data set) was used to evaluate the program and fine-tune the ML training model’s parameters. Finally, the program was presented with new images of cancer and normal cases for final assessment of accuracy (test data set). Our program learned to recognize differences from the images provided and was able to differentiate normal and cancer images with > 95% accuracy.

Advances in computer processing have allowed for the development of artificial neural networks (ANNs). While there are several types of ANNs, the most common types used for image classification and segmentation are known as convolutional neural networks (CNNs).^9,12-14 The programs are designed to work similar to the human brain, specifically the visual cortex.^15,16 As data are acquired, they are processed by various layers in the program. Much like neurons in the brain, one layer decides whether to advance information to the next.^13,14 CNNs can be many layers deep, leading to the term deep learning: “computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction.”^1,13,17

ANNs can process larger volumes of data. This advance has led to the development of unstructured or unsupervised learning. With this type of learning, imputing defined features (ie, predetermined answers) of the training data set described above is no longer required.^1,8,10,14 The advantage of unsupervised learning is that the program can be presented raw data and extract meaningful interpretation without human input, often with less bias than may exist with supervised learning.^1,18 If shown enough data, the program can extract relevant features to make conclusions independently without predefined definitions, potentially uncovering markers not previously known. For example, several studies have used unsupervised learning to search patient data to assess readmission risks of patients with congestive heart failure.^10,19,20 AI compiled features independently and not previously defined, predicting patients at greater risk for readmission superior to traditional methods.

Health Care Applications

Image analysis has seen the most AI health care applications.^8,15 AI has shown potential in interpreting many types of medical images, including pathology slides, radiographs of various types, retina and other eye scans, and photographs of skin lesions. Many studies have demonstrated that AI can interpret these images as accurately as or even better than experienced clinicians.^9,13,22-29 Studies have suggested AI interpretation of radiographs may better distinguish patients infected with COVID-19 from other causes of pneumonia, and AI interpretation of pathology slides may detect specific genetic mutations not previously identified without additional molecular tests.^{11,14,23,24,30-32}

The second area in which AI can impact health care is improving workflow and efficiency. AI has improved surgery scheduling, saving significant revenue, and decreased patient wait times for appointments.¹ AI can screen and triage radiographs, allowing attention to be directed to critical patients. This use would be valuable in many busy clinical settings, such as the recent COVID-19 pandemic.^8,23 Similarly, AI can screen retina images to prioritize urgent conditions.²⁵ AI has improved pathologists’ efficiency when used to detect breast metastases.³³ Finally, AI may reduce medical errors, thereby ensuring patient safety.^8,9,34

A third health care benefit of AI is in public health and epidemiology. AI can assist with clinical decision-making and diagnoses in low-income countries and areas with limited health care resources and personnel.^25,29 AI can improve identification of infectious outbreaks, such as tuberculosis, malaria, dengue fever, and influenza.^29,35-40 AI has been used to predict transmission patterns of the Zika virus and the current COVID-19 pandemic.^41,42 Applications can stratify the risk of outbreaks based on multiple factors, including age, income, race, atypical geographic clusters, and seasonal factors like rainfall and temperature.^35,36,38,43 AI has been used to assess morbidity and mortality, such as predicting disease severity with malaria and identifying treatment failures in tuberculosis.²⁹

Finally, AI can dramatically impact health care due to processing large data sets or disconnected volumes of patient information—so-called big data.^44-46 An example is the widespread use of electronic health records (EHRs) such as the Computerized Patient Record System used in Veteran Affairs medical centers (VAMCs). Much of patient information exists in written text: HCP notes, laboratory and radiology reports, medication records, etc. Natural language processing (NLP) allows platforms to sort through extensive volumes of data on complex patients at rates much faster than human capability, which has great potential to assist with diagnosis and treatment decisions.⁹

Medical literature is being produced at rates that exceed our ability to digest. More than 200,000 cancer-related articles were published in 2019 alone.¹⁴ NLP capabilities of AI have the potential to rapidly sort through this extensive medical literature and relate specific verbiage in patient records guiding therapy.⁴⁶ IBM Watson, a supercomputer based on ML and NLP, demonstrates this concept with many potential applications, only some of which relate to health care.^1,9 Watson has an oncology component to assimilate multiple aspects of patient care, including clinical notes, pathology results, radiograph findings, staging, and a tumor’s genetic profile. It coordinates these inputs from the EHR and mines medical literature and research databases to recommend treatment options.^1,46 AI can assess and compile far greater patient data and therapeutic options than would be feasible by individual clinicians, thus providing customized patient care.⁴⁷ Watson has partnered with numerous medical centers, including MD Anderson Cancer Center and Memorial Sloan Kettering Cancer Center, with variable success.^44,47-49 While the full potential of Watson appears not yet realized, these AI-driven approaches will likely play an important role in leveraging the hidden value in the expanding volume of health care information.

Medical Specialty Applications

Radiology

Currently > 70% of FDA-approved AI medical devices are in the field of radiology.² Most radiology departments have used AI-friendly digital imaging for years, such as the picture archiving and communication systems used by numerous health care systems, including VAMCs.^2,15 Gray-scale images common in radiology lend themselves to standardization, although AI is not limited to black-and- white image interpretation.¹⁵

An abundance of literature describes plain radiograph interpretation using AI. One FDA-approved platform improved X-ray diagnosis of wrist fractures when used by emergency medicine clinicians.^2,50 AI has been applied to chest X-ray (CXR) interpretation of many conditions, including pneumonia, tuberculosis, malignant lung lesions, and COVID-19.^{23,25,28,44,51-53} For example, Nam and colleagues suggested AI is better at diagnosing malignant pulmonary nodules from CXRs than are trained radiologists.²⁸

In addition to plain radiographs, AI has been applied to many other imaging technologies, including ultrasounds, positron emission tomography, mammograms, computed tomography (CT), and magnetic resonance imaging (MRI).^{15,26,44,48,54-56} A large study demonstrated that ML platforms significantly reduced the time to diagnose intracranial hemorrhages on CT and identified subtle hemorrhages missed by radiologists.⁵⁵ Other studies have claimed that AI programs may be better than radiologists in detecting cancer in screening mammograms, and 3 FDA-approved devices focus on mammogram interpretation.^2,15,54,57 There is also great interest in MRI applications to detect and predict prognosis for breast cancer based on imaging findings.^21,56

Aside from providing accurate diagnoses, other studies focus on AI radiograph interpretation to assist with patient screening, triage, improving time to final diagnosis, providing a rapid “second opinion,” and even monitoring disease progression and offering insights into prognosis.^{8,21,23,52,55,56,58} These features help in busy urban centers but may play an even greater role in areas with limited access to health care or trained specialists such as radiologists.⁵²

Cardiology

Cardiology has the second highest number of FDA-approved AI applications.² Many cardiology AI platforms involve image analysis, as described in several recent reviews.^45,59,60 AI has been applied to echocardiography to measure ejection fractions, detect valvular disease, and assess heart failure from hypertrophic and restrictive cardiomyopathy and amyloidosis.^45,48,59 Applications for cardiac CT scans and CT angiography have successfully quantified both calcified and noncalcified coronary artery plaques and lumen assessments, assessed myocardial perfusion, and performed coronary artery calcium scoring.^45,59,60 Likewise, AI applications for cardiac MRI have been used to quantitate ejection fraction, large vessel flow assessment, and cardiac scar burden.^45,59

For years ECG devices have provided interpretation with limited accuracy using preprogrammed parameters.⁴⁸ However, the application of AI allows ECG interpretation on par with trained cardiologists. Numerous such AI applications exist, and 2 FDA-approved devices perform ECG interpretation.^2,61-64 One of these devices incorporates an AI-powered stethoscope to detect atrial fibrillation and heart murmurs.⁶⁵

Pathology

The advancement of whole slide imaging, wherein entire slides can be scanned and digitized at high speed and resolution, creates great potential for AI applications in pathology.^{12,24,32,33,66} A landmark study demonstrating the potential of AI for assessing whole slide imaging examined sentinel lymph node metastases in patients with breast cancer.²² Multiple algorithms in the study demonstrated that AI was equivalent or better than pathologists in detecting metastases, especially when the pathologists were time-constrained consistent with a normal working environment. Significantly, the most accurate and efficient diagnoses were achieved when the pathologist and AI interpretations were used together.^22,33

AI has shown promise in diagnosing many other entities, including cancers of the prostate (including Gleason scoring), lung, colon, breast, and skin.^{11,12,24,27,32,67} In addition, AI has shown great potential in scoring biomarkers important for prognosis and treatment, such as immunohistochemistry (IHC) labeling of Ki-67 and PD-L1.³² Pathologists can have difficulty classifying certain tumors or determining the site of origin for metastases, often having to rely on IHC with limited success. The unique features of image analysis with AI have the potential to assist in classifying difficult tumors and identifying sites of origin for metastatic disease based on morphology alone.¹¹

Oncology depends heavily on molecular pathology testing to dictate treatment options and determine prognosis. Preliminary studies suggest that AI interpretation alone has the potential to delineate whether certain molecular mutations are present in tumors from various sites.^11,14,24,32 One study combined histology and genomic results for AI interpretation that improved prognostic predictions.⁶⁸ In addition, AI analysis may have potential in predicting tumor recurrence or prognosis based on cellular features, as demonstrated for lung cancer and melanoma.^67,69,70

Ophthalmology

AI applications for ophthalmology have focused on diabetic retinopathy, age-related macular degeneration, glaucoma, retinopathy of prematurity, age-related and congenital cataracts, and retinal vein occlusion.^71-73 Diabetic retinopathy is a leading cause of blindness and has been studied by numerous platforms with good success, most having used color fundus photography.^71,72 One study showed AI could diagnose diabetic retinopathy and diabetic macular edema with specificities similar to ophthalmologists.⁷⁴ In 2018, the FDA approved the AI platform IDx-DR. This diagnostic system classifies retinal images and recommends referral for patients determined to have “more than mild diabetic retinopathy” and reexamination within a year for other patients.^8,75 Significantly, the platform recommendations do not require confirmation by a clinician.⁸

AI has been applied to other modalities in ophthalmology such as optical coherence tomography (OCT) to diagnose retinal disease and to predict appropriate management of congenital cataracts.^25,73,76 For example, an AI application using OCT has been demonstrated to match or exceed the accuracy of retinal experts in diagnosing and triaging patients with a variety of retinal pathologies, including patients needing urgent referrals.⁷⁷

Dermatology

Multiple studies demonstrate AI performs at least equal to experienced dermatologists in differentiating selected skin lesions.^78-81 For example, Esteva and colleagues demonstrated AI could differentiate keratinocyte carcinomas from benign seborrheic keratoses and malignant melanomas from benign nevi with accuracy equal to 21 board-certified dermatologists.⁷⁸

AI is applicable to various imaging procedures common to dermatology, such as dermoscopy, very high-frequency ultrasound, and reflectance confocal microscopy.⁸² Several studies have demonstrated that AI interpretation compared favorably to dermatologists evaluating dermoscopy to assess melanocytic lesions.^78-81,83

A limitation in these studies is that they differentiate only a few diagnoses.⁸² Furthermore, dermatologists have sensory input such as touch and visual examination under various conditions, something AI has yet to replicate.^15,34,84 Also, most AI devices use no or limited clinical information.⁸¹ Dermatologists can recognize rarer conditions for which AI models may have had limited or no training.³⁴ Nevertheless, a recent study assessed AI for the diagnosis of 134 separate skin disorders with promising results, including providing diagnoses with accuracy comparable to that of dermatologists and providing accurate treatment strategies.⁸⁴ As Topol points out, most skin lesions are diagnosed in the primary care setting where AI can have a greater impact when used in conjunction with the clinical impression, especially where specialists are in limited supply.^48,78

Finally, dermatology lends itself to using portable or smartphone applications (apps) wherein the user can photograph a lesion for analysis by AI algorithms to assess the need for further evaluation or make treatment recommendations.^34,84,85 Although results from currently available apps are not encouraging, they may play a greater role as the technology advances.^34,85

Oncology

Applications of AI in oncology include predicting prognosis for patients with cancer based on histologic and/or genetic information.^14,68,86 Programs can predict the risk of complications before and recurrence risks after surgery for malignancies.^44,87-89 AI can also assist in treatment planning and predict treatment failure with radiation therapy.^90,91

AI has great potential in processing the large volumes of patient data in cancer genomics. Next-generation sequencing has allowed for the identification of millions of DNA sequences in a single tumor to detect genetic anomalies.⁹² Thousands of mutations can be found in individual tumor samples, and processing this information and determining its significance can be beyond human capability.¹⁴ We know little about the effects of various mutation combinations, and most tumors have a heterogeneous molecular profile among different cell populations.^14,93 The presence or absence of various mutations can have diagnostic, prognostic, and therapeutic implications.⁹³ AI has great potential to sort through these complex data and identify actionable findings.

More than 200,000 cancer-related articles were published in 2019, and publications in the field of cancer genomics are increasing exponentially.^14,92,93 Patel and colleagues assessed the utility of IBM Watson for Genomics against results from a molecular tumor board.⁹³ Watson for Genomics identified potentially significant mutations not identified by the tumor board in 32% of patients. Most mutations were related to new clinical trials not yet added to the tumor board watch list, demonstrating the role AI will have in processing the large volume of genetic data required to deliver personalized medicine moving forward.

Gastroenterology

AI has shown promise in predicting risk or outcomes based on clinical parameters in various common gastroenterology problems, including gastric reflux, acute pancreatitis, gastrointestinal bleeding, celiac disease, and inflammatory bowel disease.^94,95 AI endoscopic analysis has demonstrated potential in assessing Barrett’s esophagus, gastric Helicobacter pylori infections, gastric atrophy, and gastric intestinal metaplasia.⁹⁵ Applications have been used to assess esophageal, gastric, and colonic malignancies, including depth of invasion based on endoscopic images.⁹⁵ Finally, studies have evaluated AI to assess small colon polyps during colonoscopy, including differentiating benign and premalignant polyps with success comparable to gastroenterologists.^94,95 AI has been shown to increase the speed and accuracy of gastroenterologists in detecting small polyps during colonoscopy.⁴⁸ In a prospective randomized study, colonoscopies performed using an AI device identified significantly more small adenomatous polyps than colonoscopies without AI.⁹⁶

Neurology

It has been suggested that AI technologies are well suited for application in neurology due to the subtle presentation of many neurologic diseases.¹⁶ Viz LVO, the first CMS-approved AI reimbursement for the diagnosis of strokes, analyzes CTs to detect early ischemic strokes and alerts the medical team, thus shortening time to treatment.^3,97 Many other AI platforms are in use or development that use CT and MRI for the early detection of strokes as well as for treatment and prognosis.^9,97

AI technologies have been applied to neurodegenerative diseases, such as Alzheimer and Parkinson diseases.^16,98 For example, several studies have evaluated patient movements in Parkinson disease for both early diagnosis and to assess response to treatment.⁹⁸ These evaluations included assessment with both external cameras as well as wearable devices and smartphone apps.

Mental Health

Due to the interactive nature of mental health care, the field has been slower to develop AI applications.¹⁸ With heavy reliance on textual information (eg, clinic notes, mood rating scales, and documentation of conversations), successful AI applications in this field will likely rely heavily on NLP.¹⁸ However, studies investigating the application of AI to mental health have also incorporated data such as brain imaging, smartphone monitoring, and social media platforms, such as Facebook and Twitter.^18,103,104

The risk of suicide is higher in veteran patients, and ML algorithms have had limited success in predicting suicide risk in both veteran and nonveteran populations.^104-106 While early models have low positive predictive values and low sensitivities, they still promise to be a useful tool in conjunction with traditional risk assessments.¹⁰⁶ Kessler and colleagues suggest that combining multiple rather than single ML algorithms might lead to greater success.^105,106

AI may assist in diagnosing other mental health disorders, including major depressive disorder, attention deficit hyperactivity disorder (ADHD), schizophrenia, posttraumatic stress disorder, and Alzheimer disease.^103,104,107 These investigations are in the early stages with limited clinical applicability. However, 2 AI applications awaiting FDA approval relate to ADHD and opioid use.² Furthermore, potential exists for AI to not only assist with prevention and diagnosis of ADHD, but also to identify optimal treatment options.^2,103

General and Personalized Medicine

Additional AI applications include diagnosing patients with suspected sepsis, measuring liver iron concentrations, predicting hospital mortality at the time of admission, and more.^2,108,109 AI can guide end-of-life decisions such as resuscitation status or whether to initiate mechanical ventilation.⁴⁸

AI-driven smartphone apps can be beneficial to both patients and clinicians. Examples include predicting nonadherence to anticoagulation therapy, monitoring heart rhythms for atrial fibrillation or signs of hyperkalemia in patients with renal failure, and improving outcomes for patients with diabetes mellitus by decreasing glycemic variability and reducing hypoglycemia.^8,48,110,111 The potential for AI applications to health care and personalized medicine are almost limitless.

Discussion

With ever-increasing expectations for all health care sectors to deliver timely, fiscally-responsible, high-quality health care, AI has the potential to have numerous impacts. AI can improve diagnostic accuracy while limiting errors and impact patient safety such as assisting with prescription delivery.^8,9,34 It can screen and triage patients, alerting clinicians to those needing more urgent evaluation.^8,23,77,97 AI also may increase a clinician’s efficiency and speed to render a diagnosis.^12,13,55,97 AI can provide a rapid second opinion, an ability especially beneficial in underserved areas with shortages of specialists.^{23,25,26,29,34} Similarly, AI may decrease the inter- and intraobserver variability common in many medical specialties.^12,27,45 AI applications can also monitor disease progression, identifying patients at greatest risk, and provide information for prognosis.^21,23,56,58 Finally, as described with applications using IBM Watson, AI can allow for an integrated approach to health care that is currently lacking.

We have described many reports suggesting AI can render diagnoses as well as or better than experienced clinicians, and speculation exists that AI will replace many roles currently performed by health care practitioners.^9,26 However, most studies demonstrate that AI’s diagnostic benefits are best realized when used to supplement a clinician’s impression.^{8,22,30,33,52,54,56,69,84} AI is not likely to replace humans in health care in the foreseeable future. The technology can be likened to the impact of CT scans developed in the 1970s in neurology. Prior to such detailed imaging, neurologists spent extensive time performing detailed physicals to render diagnoses and locate lesions before surgery. There was mistrust of this new technology and concern that CT scans would eliminate the need for neurologists.¹¹² On the contrary, neurology is alive and well, frequently being augmented by the technologies once speculated to replace it.

Commercial AI health care platforms represented a $2 billion industry in 2018 and are growing rapidly each year.^13,32 Many AI products are offered ready for implementation for various tasks, including diagnostics, patient management, and improved efficiency. Others will likely be provided as templates suitable for modification to meet the specific needs of the facility, practice, or specialty for its patient population.

AI Risks and Limitations

AI has several risks and limitations. Although there is progress in explainable AI, at times we still struggle to understand how the output provided by machine learning algorithms was created.^44,48 The many layers associated with deep learning self-determine the criteria to reach its conclusion, and these criteria can continually evolve. The parameters of deep learning are not preprogrammed, and there are too many individual data points to be extrapolated or deconvoluted for evaluation at our current level of knowledge.^26,51 These apparent lack of constraints cause concern for patient safety and suggest that greater validation and continued scrutiny of validity is required.^8,48 Efforts are underway to create explainable AI programs to make their processes more transparent, but such clarification is limited presently.^14,26,48,77

Another challenge of AI is determining the amount of training data required to function optimally. Also, if the output describes multiple variables or diagnoses, are each equally valid?¹¹³ Furthermore, many AI applications look for a specific process, such as cancer diagnoses on CXRs. However, how coexisting conditions like cardiomegaly, emphysema, pneumonia, etc, seen on CXRs will affect the diagnosis needs to be considered.^51,52 Zech and colleagues provide the example that diagnoses for pneumothorax are frequently rendered on CXRs with chest tubes in place.⁵¹ They suggest that CNNs may develop a bias toward diagnosing pneumothorax when chest tubes are present. Many current studies approach an issue in isolation, a situation not realistic in real-world clinical practice.²⁶

Most studies on AI have been retrospective, and frequently data used to train the program are preselected.^13,26 The data are typically validated on available databases rather than actual patients in the clinical setting, limiting confidence in the validity of the AI output when applied to real-world situations. Currently, fewer than 12 prospective trials had been published comparing AI with traditional clinical care.^13,114 Randomized prospective clinical trials are even fewer, with none currently reported from the United States.^13,114 The results from several studies have been shown to diminish when repeated prospectively.¹¹⁴

The FDA has created a new category known as Software as a Medical Device and has a Digital Health Innovation Action Plan to regulate AI platforms. Still, the process of AI regulation is of necessity different from traditional approval processes and is continually evolving.⁸ The FDA approval process cannot account for the fact that the program’s parameters may continually evolve or adapt.²

Guidelines for investigating and reporting AI research with its unique attributes are being developed. Examples include the TRIPOD-ML statement and others.^49,115 In September 2020, 2 publications addressed the paucity of gold-standard randomized clinical trials in clinical AI applications.^116,117 The SPIRIT-AI statement expands on the original SPIRIT statement published in 2013 to guide minimal reporting standards for AI clinical trial protocols to promote transparency of design and methodology.¹¹⁶ Similarly, the CONSORT-AI extension, stemming from the original CONSORT statement in 1996, aims to ensure quality reporting of randomized controlled trials in AI.¹¹⁷

Another risk with AI is that while an individual physician making a mistake may adversely affect 1 patient, a single mistake in an AI algorithm could potentially affect thousands of patients.⁴⁸ Also, AI programs developed for patient populations at a facility may not translate to another. Referred to as overfitting, this phenomenon relates to selection bias in training data sets.^{15,34,49,51,52} Studies have shown that programs that underrepresent certain group characteristics such as age, sex, or race may be less effective when applied to a population in which these characteristics have differing representations.^8,48,49 This problem of underrepresentation has been demonstrated in programs interpreting pathology slides, radiographs, and skin lesions.^15,32,51

Admittedly, most of these challenges are not specific to AI and existed in health care previously. Physicians make mistakes, treatments are sometimes used without adequate prospective studies, and medications are given without understanding their mechanism of action, much like AI-facilitated processes reach a conclusion that cannot be fully explained.⁴⁸

Conclusions

The view that AI will dramatically impact health care in the coming years will likely prove true. However, much work is needed, especially because of the paucity of prospective clinical trials as has been historically required in medical research. Any concern that AI will replace HCPs seems unwarranted. Early studies suggest that even AI programs that appear to exceed human interpretation perform best when working in cooperation with and oversight from clinicians. AI’s greatest potential appears to be its ability to augment care from health professionals, improving efficiency and accuracy, and should be anticipated with enthusiasm as the field moves forward at an exponential rate.

Acknowledgments

The authors thank Makenna G. Thomas for proofreading and review of the manuscript. This material is the result of work supported with resources and the use of facilities at the James A. Haley Veterans’ Hospital. This research has been approved by the James A. Haley Veteran’s Hospital Office of Communications and Media.

Artificial Intelligence (AI) was first described in 1956 and refers to machines having the ability to learn as they receive and process information, resulting in the ability to “think” like humans.¹ AI’s impact in medicine is increasing; currently, at least 29 AI medical devices and algorithms are approved by the US Food and Drug Administration (FDA) in a variety of areas, including radiograph interpretation, managing glucose levels in patients with diabetes mellitus, analyzing electrocardiograms (ECGs), and diagnosing sleep disorders among others.² Significantly, in 2020, the Centers for Medicare and Medicaid Services (CMS) announced the first reimbursement to hospitals for an AI platform, a model for early detection of strokes.³ AI is rapidly becoming an integral part of health care, and its role will only increase in the future (Table).

As knowledge in medicine is expanding exponentially, AI has great potential to assist with handling complex patient care data. The concept of exponential growth is not a natural one. As Bini described, with exponential growth the volume of knowledge amassed over the past 10 years will now occur in perhaps only 1 year.¹ Likewise, equivalent advances over the past year may take just a few months. This phenomenon is partly due to the law of accelerating returns, which states that advances feed on themselves, continually increasing the rate of further advances.⁴ The volume of medical data doubles every 2 to 5 years.⁵ Fortunately, the field of AI is growing exponentially as well and can help health care practitioners (HCPs) keep pace, allowing the continued delivery of effective health care.

In this report, we review common terminology, principles, and general applications of AI, followed by current and potential applications of AI for selected medical specialties. Finally, we discuss AI’s future in health care, along with potential risks and pitfalls.

AI Overview

AI refers to machine programs that can “learn” or think based on past experiences. This functionality contrasts with simple rules-based programming available to health care for years. An example of rules-based programming is the warfarindosing.org website developed by Barnes-Jewish Hospital at Washington University Medical Center, which guides initial warfarin dosing.^6,7 The prescriber inputs detailed patient information, including age, sex, height, weight, tobacco history, medications, laboratory results, and genotype if available. The application then calculates recommended warfarin dosing regimens to avoid over- or underanticoagulation. While the dosing algorithm may be complex, it depends entirely on preprogrammed rules. The program does not learn to reach its conclusions and recommendations from patient data.

In contrast, one of the most common subsets of AI is machine learning (ML). ML describes a program that “learns from experience and improves its performance as it learns.”¹ With ML, the computer is initially provided with a training data set—data with known outcomes or labels. Because the initial data are input from known samples, this type of AI is known as supervised learning.^8-10 As an example, we recently reported using ML to diagnose various types of cancer from pathology slides.¹¹ In one experiment, we captured images of colon adenocarcinoma and normal colon (these 2 groups represent the training data set). Unlike traditional programming, we did not define characteristics that would differentiate colon cancer from normal; rather, the machine learned these characteristics independently by assessing the labeled images provided. A second data set (the validation data set) was used to evaluate the program and fine-tune the ML training model’s parameters. Finally, the program was presented with new images of cancer and normal cases for final assessment of accuracy (test data set). Our program learned to recognize differences from the images provided and was able to differentiate normal and cancer images with > 95% accuracy.

Advances in computer processing have allowed for the development of artificial neural networks (ANNs). While there are several types of ANNs, the most common types used for image classification and segmentation are known as convolutional neural networks (CNNs).^9,12-14 The programs are designed to work similar to the human brain, specifically the visual cortex.^15,16 As data are acquired, they are processed by various layers in the program. Much like neurons in the brain, one layer decides whether to advance information to the next.^13,14 CNNs can be many layers deep, leading to the term deep learning: “computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction.”^1,13,17

ANNs can process larger volumes of data. This advance has led to the development of unstructured or unsupervised learning. With this type of learning, imputing defined features (ie, predetermined answers) of the training data set described above is no longer required.^1,8,10,14 The advantage of unsupervised learning is that the program can be presented raw data and extract meaningful interpretation without human input, often with less bias than may exist with supervised learning.^1,18 If shown enough data, the program can extract relevant features to make conclusions independently without predefined definitions, potentially uncovering markers not previously known. For example, several studies have used unsupervised learning to search patient data to assess readmission risks of patients with congestive heart failure.^10,19,20 AI compiled features independently and not previously defined, predicting patients at greater risk for readmission superior to traditional methods.

Health Care Applications

Image analysis has seen the most AI health care applications.^8,15 AI has shown potential in interpreting many types of medical images, including pathology slides, radiographs of various types, retina and other eye scans, and photographs of skin lesions. Many studies have demonstrated that AI can interpret these images as accurately as or even better than experienced clinicians.^9,13,22-29 Studies have suggested AI interpretation of radiographs may better distinguish patients infected with COVID-19 from other causes of pneumonia, and AI interpretation of pathology slides may detect specific genetic mutations not previously identified without additional molecular tests.^{11,14,23,24,30-32}

The second area in which AI can impact health care is improving workflow and efficiency. AI has improved surgery scheduling, saving significant revenue, and decreased patient wait times for appointments.¹ AI can screen and triage radiographs, allowing attention to be directed to critical patients. This use would be valuable in many busy clinical settings, such as the recent COVID-19 pandemic.^8,23 Similarly, AI can screen retina images to prioritize urgent conditions.²⁵ AI has improved pathologists’ efficiency when used to detect breast metastases.³³ Finally, AI may reduce medical errors, thereby ensuring patient safety.^8,9,34

A third health care benefit of AI is in public health and epidemiology. AI can assist with clinical decision-making and diagnoses in low-income countries and areas with limited health care resources and personnel.^25,29 AI can improve identification of infectious outbreaks, such as tuberculosis, malaria, dengue fever, and influenza.^29,35-40 AI has been used to predict transmission patterns of the Zika virus and the current COVID-19 pandemic.^41,42 Applications can stratify the risk of outbreaks based on multiple factors, including age, income, race, atypical geographic clusters, and seasonal factors like rainfall and temperature.^35,36,38,43 AI has been used to assess morbidity and mortality, such as predicting disease severity with malaria and identifying treatment failures in tuberculosis.²⁹

Finally, AI can dramatically impact health care due to processing large data sets or disconnected volumes of patient information—so-called big data.^44-46 An example is the widespread use of electronic health records (EHRs) such as the Computerized Patient Record System used in Veteran Affairs medical centers (VAMCs). Much of patient information exists in written text: HCP notes, laboratory and radiology reports, medication records, etc. Natural language processing (NLP) allows platforms to sort through extensive volumes of data on complex patients at rates much faster than human capability, which has great potential to assist with diagnosis and treatment decisions.⁹

Medical literature is being produced at rates that exceed our ability to digest. More than 200,000 cancer-related articles were published in 2019 alone.¹⁴ NLP capabilities of AI have the potential to rapidly sort through this extensive medical literature and relate specific verbiage in patient records guiding therapy.⁴⁶ IBM Watson, a supercomputer based on ML and NLP, demonstrates this concept with many potential applications, only some of which relate to health care.^1,9 Watson has an oncology component to assimilate multiple aspects of patient care, including clinical notes, pathology results, radiograph findings, staging, and a tumor’s genetic profile. It coordinates these inputs from the EHR and mines medical literature and research databases to recommend treatment options.^1,46 AI can assess and compile far greater patient data and therapeutic options than would be feasible by individual clinicians, thus providing customized patient care.⁴⁷ Watson has partnered with numerous medical centers, including MD Anderson Cancer Center and Memorial Sloan Kettering Cancer Center, with variable success.^44,47-49 While the full potential of Watson appears not yet realized, these AI-driven approaches will likely play an important role in leveraging the hidden value in the expanding volume of health care information.

Medical Specialty Applications

Radiology

Currently > 70% of FDA-approved AI medical devices are in the field of radiology.² Most radiology departments have used AI-friendly digital imaging for years, such as the picture archiving and communication systems used by numerous health care systems, including VAMCs.^2,15 Gray-scale images common in radiology lend themselves to standardization, although AI is not limited to black-and- white image interpretation.¹⁵

An abundance of literature describes plain radiograph interpretation using AI. One FDA-approved platform improved X-ray diagnosis of wrist fractures when used by emergency medicine clinicians.^2,50 AI has been applied to chest X-ray (CXR) interpretation of many conditions, including pneumonia, tuberculosis, malignant lung lesions, and COVID-19.^{23,25,28,44,51-53} For example, Nam and colleagues suggested AI is better at diagnosing malignant pulmonary nodules from CXRs than are trained radiologists.²⁸

In addition to plain radiographs, AI has been applied to many other imaging technologies, including ultrasounds, positron emission tomography, mammograms, computed tomography (CT), and magnetic resonance imaging (MRI).^{15,26,44,48,54-56} A large study demonstrated that ML platforms significantly reduced the time to diagnose intracranial hemorrhages on CT and identified subtle hemorrhages missed by radiologists.⁵⁵ Other studies have claimed that AI programs may be better than radiologists in detecting cancer in screening mammograms, and 3 FDA-approved devices focus on mammogram interpretation.^2,15,54,57 There is also great interest in MRI applications to detect and predict prognosis for breast cancer based on imaging findings.^21,56

Aside from providing accurate diagnoses, other studies focus on AI radiograph interpretation to assist with patient screening, triage, improving time to final diagnosis, providing a rapid “second opinion,” and even monitoring disease progression and offering insights into prognosis.^{8,21,23,52,55,56,58} These features help in busy urban centers but may play an even greater role in areas with limited access to health care or trained specialists such as radiologists.⁵²

Cardiology

Cardiology has the second highest number of FDA-approved AI applications.² Many cardiology AI platforms involve image analysis, as described in several recent reviews.^45,59,60 AI has been applied to echocardiography to measure ejection fractions, detect valvular disease, and assess heart failure from hypertrophic and restrictive cardiomyopathy and amyloidosis.^45,48,59 Applications for cardiac CT scans and CT angiography have successfully quantified both calcified and noncalcified coronary artery plaques and lumen assessments, assessed myocardial perfusion, and performed coronary artery calcium scoring.^45,59,60 Likewise, AI applications for cardiac MRI have been used to quantitate ejection fraction, large vessel flow assessment, and cardiac scar burden.^45,59

For years ECG devices have provided interpretation with limited accuracy using preprogrammed parameters.⁴⁸ However, the application of AI allows ECG interpretation on par with trained cardiologists. Numerous such AI applications exist, and 2 FDA-approved devices perform ECG interpretation.^2,61-64 One of these devices incorporates an AI-powered stethoscope to detect atrial fibrillation and heart murmurs.⁶⁵

Pathology

The advancement of whole slide imaging, wherein entire slides can be scanned and digitized at high speed and resolution, creates great potential for AI applications in pathology.^{12,24,32,33,66} A landmark study demonstrating the potential of AI for assessing whole slide imaging examined sentinel lymph node metastases in patients with breast cancer.²² Multiple algorithms in the study demonstrated that AI was equivalent or better than pathologists in detecting metastases, especially when the pathologists were time-constrained consistent with a normal working environment. Significantly, the most accurate and efficient diagnoses were achieved when the pathologist and AI interpretations were used together.^22,33

AI has shown promise in diagnosing many other entities, including cancers of the prostate (including Gleason scoring), lung, colon, breast, and skin.^{11,12,24,27,32,67} In addition, AI has shown great potential in scoring biomarkers important for prognosis and treatment, such as immunohistochemistry (IHC) labeling of Ki-67 and PD-L1.³² Pathologists can have difficulty classifying certain tumors or determining the site of origin for metastases, often having to rely on IHC with limited success. The unique features of image analysis with AI have the potential to assist in classifying difficult tumors and identifying sites of origin for metastatic disease based on morphology alone.¹¹

Oncology depends heavily on molecular pathology testing to dictate treatment options and determine prognosis. Preliminary studies suggest that AI interpretation alone has the potential to delineate whether certain molecular mutations are present in tumors from various sites.^11,14,24,32 One study combined histology and genomic results for AI interpretation that improved prognostic predictions.⁶⁸ In addition, AI analysis may have potential in predicting tumor recurrence or prognosis based on cellular features, as demonstrated for lung cancer and melanoma.^67,69,70

Ophthalmology

AI applications for ophthalmology have focused on diabetic retinopathy, age-related macular degeneration, glaucoma, retinopathy of prematurity, age-related and congenital cataracts, and retinal vein occlusion.^71-73 Diabetic retinopathy is a leading cause of blindness and has been studied by numerous platforms with good success, most having used color fundus photography.^71,72 One study showed AI could diagnose diabetic retinopathy and diabetic macular edema with specificities similar to ophthalmologists.⁷⁴ In 2018, the FDA approved the AI platform IDx-DR. This diagnostic system classifies retinal images and recommends referral for patients determined to have “more than mild diabetic retinopathy” and reexamination within a year for other patients.^8,75 Significantly, the platform recommendations do not require confirmation by a clinician.⁸

AI has been applied to other modalities in ophthalmology such as optical coherence tomography (OCT) to diagnose retinal disease and to predict appropriate management of congenital cataracts.^25,73,76 For example, an AI application using OCT has been demonstrated to match or exceed the accuracy of retinal experts in diagnosing and triaging patients with a variety of retinal pathologies, including patients needing urgent referrals.⁷⁷

Dermatology

Multiple studies demonstrate AI performs at least equal to experienced dermatologists in differentiating selected skin lesions.^78-81 For example, Esteva and colleagues demonstrated AI could differentiate keratinocyte carcinomas from benign seborrheic keratoses and malignant melanomas from benign nevi with accuracy equal to 21 board-certified dermatologists.⁷⁸

AI is applicable to various imaging procedures common to dermatology, such as dermoscopy, very high-frequency ultrasound, and reflectance confocal microscopy.⁸² Several studies have demonstrated that AI interpretation compared favorably to dermatologists evaluating dermoscopy to assess melanocytic lesions.^78-81,83

A limitation in these studies is that they differentiate only a few diagnoses.⁸² Furthermore, dermatologists have sensory input such as touch and visual examination under various conditions, something AI has yet to replicate.^15,34,84 Also, most AI devices use no or limited clinical information.⁸¹ Dermatologists can recognize rarer conditions for which AI models may have had limited or no training.³⁴ Nevertheless, a recent study assessed AI for the diagnosis of 134 separate skin disorders with promising results, including providing diagnoses with accuracy comparable to that of dermatologists and providing accurate treatment strategies.⁸⁴ As Topol points out, most skin lesions are diagnosed in the primary care setting where AI can have a greater impact when used in conjunction with the clinical impression, especially where specialists are in limited supply.^48,78

Finally, dermatology lends itself to using portable or smartphone applications (apps) wherein the user can photograph a lesion for analysis by AI algorithms to assess the need for further evaluation or make treatment recommendations.^34,84,85 Although results from currently available apps are not encouraging, they may play a greater role as the technology advances.^34,85

Oncology

Applications of AI in oncology include predicting prognosis for patients with cancer based on histologic and/or genetic information.^14,68,86 Programs can predict the risk of complications before and recurrence risks after surgery for malignancies.^44,87-89 AI can also assist in treatment planning and predict treatment failure with radiation therapy.^90,91

AI has great potential in processing the large volumes of patient data in cancer genomics. Next-generation sequencing has allowed for the identification of millions of DNA sequences in a single tumor to detect genetic anomalies.⁹² Thousands of mutations can be found in individual tumor samples, and processing this information and determining its significance can be beyond human capability.¹⁴ We know little about the effects of various mutation combinations, and most tumors have a heterogeneous molecular profile among different cell populations.^14,93 The presence or absence of various mutations can have diagnostic, prognostic, and therapeutic implications.⁹³ AI has great potential to sort through these complex data and identify actionable findings.

More than 200,000 cancer-related articles were published in 2019, and publications in the field of cancer genomics are increasing exponentially.^14,92,93 Patel and colleagues assessed the utility of IBM Watson for Genomics against results from a molecular tumor board.⁹³ Watson for Genomics identified potentially significant mutations not identified by the tumor board in 32% of patients. Most mutations were related to new clinical trials not yet added to the tumor board watch list, demonstrating the role AI will have in processing the large volume of genetic data required to deliver personalized medicine moving forward.

Gastroenterology

AI has shown promise in predicting risk or outcomes based on clinical parameters in various common gastroenterology problems, including gastric reflux, acute pancreatitis, gastrointestinal bleeding, celiac disease, and inflammatory bowel disease.^94,95 AI endoscopic analysis has demonstrated potential in assessing Barrett’s esophagus, gastric Helicobacter pylori infections, gastric atrophy, and gastric intestinal metaplasia.⁹⁵ Applications have been used to assess esophageal, gastric, and colonic malignancies, including depth of invasion based on endoscopic images.⁹⁵ Finally, studies have evaluated AI to assess small colon polyps during colonoscopy, including differentiating benign and premalignant polyps with success comparable to gastroenterologists.^94,95 AI has been shown to increase the speed and accuracy of gastroenterologists in detecting small polyps during colonoscopy.⁴⁸ In a prospective randomized study, colonoscopies performed using an AI device identified significantly more small adenomatous polyps than colonoscopies without AI.⁹⁶

Neurology

It has been suggested that AI technologies are well suited for application in neurology due to the subtle presentation of many neurologic diseases.¹⁶ Viz LVO, the first CMS-approved AI reimbursement for the diagnosis of strokes, analyzes CTs to detect early ischemic strokes and alerts the medical team, thus shortening time to treatment.^3,97 Many other AI platforms are in use or development that use CT and MRI for the early detection of strokes as well as for treatment and prognosis.^9,97

AI technologies have been applied to neurodegenerative diseases, such as Alzheimer and Parkinson diseases.^16,98 For example, several studies have evaluated patient movements in Parkinson disease for both early diagnosis and to assess response to treatment.⁹⁸ These evaluations included assessment with both external cameras as well as wearable devices and smartphone apps.

Mental Health

Due to the interactive nature of mental health care, the field has been slower to develop AI applications.¹⁸ With heavy reliance on textual information (eg, clinic notes, mood rating scales, and documentation of conversations), successful AI applications in this field will likely rely heavily on NLP.¹⁸ However, studies investigating the application of AI to mental health have also incorporated data such as brain imaging, smartphone monitoring, and social media platforms, such as Facebook and Twitter.^18,103,104

The risk of suicide is higher in veteran patients, and ML algorithms have had limited success in predicting suicide risk in both veteran and nonveteran populations.^104-106 While early models have low positive predictive values and low sensitivities, they still promise to be a useful tool in conjunction with traditional risk assessments.¹⁰⁶ Kessler and colleagues suggest that combining multiple rather than single ML algorithms might lead to greater success.^105,106

AI may assist in diagnosing other mental health disorders, including major depressive disorder, attention deficit hyperactivity disorder (ADHD), schizophrenia, posttraumatic stress disorder, and Alzheimer disease.^103,104,107 These investigations are in the early stages with limited clinical applicability. However, 2 AI applications awaiting FDA approval relate to ADHD and opioid use.² Furthermore, potential exists for AI to not only assist with prevention and diagnosis of ADHD, but also to identify optimal treatment options.^2,103

General and Personalized Medicine

Additional AI applications include diagnosing patients with suspected sepsis, measuring liver iron concentrations, predicting hospital mortality at the time of admission, and more.^2,108,109 AI can guide end-of-life decisions such as resuscitation status or whether to initiate mechanical ventilation.⁴⁸

AI-driven smartphone apps can be beneficial to both patients and clinicians. Examples include predicting nonadherence to anticoagulation therapy, monitoring heart rhythms for atrial fibrillation or signs of hyperkalemia in patients with renal failure, and improving outcomes for patients with diabetes mellitus by decreasing glycemic variability and reducing hypoglycemia.^8,48,110,111 The potential for AI applications to health care and personalized medicine are almost limitless.

Discussion

With ever-increasing expectations for all health care sectors to deliver timely, fiscally-responsible, high-quality health care, AI has the potential to have numerous impacts. AI can improve diagnostic accuracy while limiting errors and impact patient safety such as assisting with prescription delivery.^8,9,34 It can screen and triage patients, alerting clinicians to those needing more urgent evaluation.^8,23,77,97 AI also may increase a clinician’s efficiency and speed to render a diagnosis.^12,13,55,97 AI can provide a rapid second opinion, an ability especially beneficial in underserved areas with shortages of specialists.^{23,25,26,29,34} Similarly, AI may decrease the inter- and intraobserver variability common in many medical specialties.^12,27,45 AI applications can also monitor disease progression, identifying patients at greatest risk, and provide information for prognosis.^21,23,56,58 Finally, as described with applications using IBM Watson, AI can allow for an integrated approach to health care that is currently lacking.

We have described many reports suggesting AI can render diagnoses as well as or better than experienced clinicians, and speculation exists that AI will replace many roles currently performed by health care practitioners.^9,26 However, most studies demonstrate that AI’s diagnostic benefits are best realized when used to supplement a clinician’s impression.^{8,22,30,33,52,54,56,69,84} AI is not likely to replace humans in health care in the foreseeable future. The technology can be likened to the impact of CT scans developed in the 1970s in neurology. Prior to such detailed imaging, neurologists spent extensive time performing detailed physicals to render diagnoses and locate lesions before surgery. There was mistrust of this new technology and concern that CT scans would eliminate the need for neurologists.¹¹² On the contrary, neurology is alive and well, frequently being augmented by the technologies once speculated to replace it.

Commercial AI health care platforms represented a $2 billion industry in 2018 and are growing rapidly each year.^13,32 Many AI products are offered ready for implementation for various tasks, including diagnostics, patient management, and improved efficiency. Others will likely be provided as templates suitable for modification to meet the specific needs of the facility, practice, or specialty for its patient population.

AI Risks and Limitations

AI has several risks and limitations. Although there is progress in explainable AI, at times we still struggle to understand how the output provided by machine learning algorithms was created.^44,48 The many layers associated with deep learning self-determine the criteria to reach its conclusion, and these criteria can continually evolve. The parameters of deep learning are not preprogrammed, and there are too many individual data points to be extrapolated or deconvoluted for evaluation at our current level of knowledge.^26,51 These apparent lack of constraints cause concern for patient safety and suggest that greater validation and continued scrutiny of validity is required.^8,48 Efforts are underway to create explainable AI programs to make their processes more transparent, but such clarification is limited presently.^14,26,48,77

Another challenge of AI is determining the amount of training data required to function optimally. Also, if the output describes multiple variables or diagnoses, are each equally valid?¹¹³ Furthermore, many AI applications look for a specific process, such as cancer diagnoses on CXRs. However, how coexisting conditions like cardiomegaly, emphysema, pneumonia, etc, seen on CXRs will affect the diagnosis needs to be considered.^51,52 Zech and colleagues provide the example that diagnoses for pneumothorax are frequently rendered on CXRs with chest tubes in place.⁵¹ They suggest that CNNs may develop a bias toward diagnosing pneumothorax when chest tubes are present. Many current studies approach an issue in isolation, a situation not realistic in real-world clinical practice.²⁶

Most studies on AI have been retrospective, and frequently data used to train the program are preselected.^13,26 The data are typically validated on available databases rather than actual patients in the clinical setting, limiting confidence in the validity of the AI output when applied to real-world situations. Currently, fewer than 12 prospective trials had been published comparing AI with traditional clinical care.^13,114 Randomized prospective clinical trials are even fewer, with none currently reported from the United States.^13,114 The results from several studies have been shown to diminish when repeated prospectively.¹¹⁴

The FDA has created a new category known as Software as a Medical Device and has a Digital Health Innovation Action Plan to regulate AI platforms. Still, the process of AI regulation is of necessity different from traditional approval processes and is continually evolving.⁸ The FDA approval process cannot account for the fact that the program’s parameters may continually evolve or adapt.²

Guidelines for investigating and reporting AI research with its unique attributes are being developed. Examples include the TRIPOD-ML statement and others.^49,115 In September 2020, 2 publications addressed the paucity of gold-standard randomized clinical trials in clinical AI applications.^116,117 The SPIRIT-AI statement expands on the original SPIRIT statement published in 2013 to guide minimal reporting standards for AI clinical trial protocols to promote transparency of design and methodology.¹¹⁶ Similarly, the CONSORT-AI extension, stemming from the original CONSORT statement in 1996, aims to ensure quality reporting of randomized controlled trials in AI.¹¹⁷

Another risk with AI is that while an individual physician making a mistake may adversely affect 1 patient, a single mistake in an AI algorithm could potentially affect thousands of patients.⁴⁸ Also, AI programs developed for patient populations at a facility may not translate to another. Referred to as overfitting, this phenomenon relates to selection bias in training data sets.^{15,34,49,51,52} Studies have shown that programs that underrepresent certain group characteristics such as age, sex, or race may be less effective when applied to a population in which these characteristics have differing representations.^8,48,49 This problem of underrepresentation has been demonstrated in programs interpreting pathology slides, radiographs, and skin lesions.^15,32,51

Admittedly, most of these challenges are not specific to AI and existed in health care previously. Physicians make mistakes, treatments are sometimes used without adequate prospective studies, and medications are given without understanding their mechanism of action, much like AI-facilitated processes reach a conclusion that cannot be fully explained.⁴⁸

Conclusions

The view that AI will dramatically impact health care in the coming years will likely prove true. However, much work is needed, especially because of the paucity of prospective clinical trials as has been historically required in medical research. Any concern that AI will replace HCPs seems unwarranted. Early studies suggest that even AI programs that appear to exceed human interpretation perform best when working in cooperation with and oversight from clinicians. AI’s greatest potential appears to be its ability to augment care from health professionals, improving efficiency and accuracy, and should be anticipated with enthusiasm as the field moves forward at an exponential rate.

Acknowledgments

The authors thank Makenna G. Thomas for proofreading and review of the manuscript. This material is the result of work supported with resources and the use of facilities at the James A. Haley Veterans’ Hospital. This research has been approved by the James A. Haley Veteran’s Hospital Office of Communications and Media.

Artificial Intelligence (AI) was first described in 1956 and refers to machines having the ability to learn as they receive and process information, resulting in the ability to “think” like humans.¹ AI’s impact in medicine is increasing; currently, at least 29 AI medical devices and algorithms are approved by the US Food and Drug Administration (FDA) in a variety of areas, including radiograph interpretation, managing glucose levels in patients with diabetes mellitus, analyzing electrocardiograms (ECGs), and diagnosing sleep disorders among others.² Significantly, in 2020, the Centers for Medicare and Medicaid Services (CMS) announced the first reimbursement to hospitals for an AI platform, a model for early detection of strokes.³ AI is rapidly becoming an integral part of health care, and its role will only increase in the future (Table).

As knowledge in medicine is expanding exponentially, AI has great potential to assist with handling complex patient care data. The concept of exponential growth is not a natural one. As Bini described, with exponential growth the volume of knowledge amassed over the past 10 years will now occur in perhaps only 1 year.¹ Likewise, equivalent advances over the past year may take just a few months. This phenomenon is partly due to the law of accelerating returns, which states that advances feed on themselves, continually increasing the rate of further advances.⁴ The volume of medical data doubles every 2 to 5 years.⁵ Fortunately, the field of AI is growing exponentially as well and can help health care practitioners (HCPs) keep pace, allowing the continued delivery of effective health care.

In this report, we review common terminology, principles, and general applications of AI, followed by current and potential applications of AI for selected medical specialties. Finally, we discuss AI’s future in health care, along with potential risks and pitfalls.

AI Overview

AI refers to machine programs that can “learn” or think based on past experiences. This functionality contrasts with simple rules-based programming available to health care for years. An example of rules-based programming is the warfarindosing.org website developed by Barnes-Jewish Hospital at Washington University Medical Center, which guides initial warfarin dosing.^6,7 The prescriber inputs detailed patient information, including age, sex, height, weight, tobacco history, medications, laboratory results, and genotype if available. The application then calculates recommended warfarin dosing regimens to avoid over- or underanticoagulation. While the dosing algorithm may be complex, it depends entirely on preprogrammed rules. The program does not learn to reach its conclusions and recommendations from patient data.

In contrast, one of the most common subsets of AI is machine learning (ML). ML describes a program that “learns from experience and improves its performance as it learns.”¹ With ML, the computer is initially provided with a training data set—data with known outcomes or labels. Because the initial data are input from known samples, this type of AI is known as supervised learning.^8-10 As an example, we recently reported using ML to diagnose various types of cancer from pathology slides.¹¹ In one experiment, we captured images of colon adenocarcinoma and normal colon (these 2 groups represent the training data set). Unlike traditional programming, we did not define characteristics that would differentiate colon cancer from normal; rather, the machine learned these characteristics independently by assessing the labeled images provided. A second data set (the validation data set) was used to evaluate the program and fine-tune the ML training model’s parameters. Finally, the program was presented with new images of cancer and normal cases for final assessment of accuracy (test data set). Our program learned to recognize differences from the images provided and was able to differentiate normal and cancer images with > 95% accuracy.

Advances in computer processing have allowed for the development of artificial neural networks (ANNs). While there are several types of ANNs, the most common types used for image classification and segmentation are known as convolutional neural networks (CNNs).^9,12-14 The programs are designed to work similar to the human brain, specifically the visual cortex.^15,16 As data are acquired, they are processed by various layers in the program. Much like neurons in the brain, one layer decides whether to advance information to the next.^13,14 CNNs can be many layers deep, leading to the term deep learning: “computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction.”^1,13,17

ANNs can process larger volumes of data. This advance has led to the development of unstructured or unsupervised learning. With this type of learning, imputing defined features (ie, predetermined answers) of the training data set described above is no longer required.^1,8,10,14 The advantage of unsupervised learning is that the program can be presented raw data and extract meaningful interpretation without human input, often with less bias than may exist with supervised learning.^1,18 If shown enough data, the program can extract relevant features to make conclusions independently without predefined definitions, potentially uncovering markers not previously known. For example, several studies have used unsupervised learning to search patient data to assess readmission risks of patients with congestive heart failure.^10,19,20 AI compiled features independently and not previously defined, predicting patients at greater risk for readmission superior to traditional methods.

Health Care Applications

Image analysis has seen the most AI health care applications.^8,15 AI has shown potential in interpreting many types of medical images, including pathology slides, radiographs of various types, retina and other eye scans, and photographs of skin lesions. Many studies have demonstrated that AI can interpret these images as accurately as or even better than experienced clinicians.^9,13,22-29 Studies have suggested AI interpretation of radiographs may better distinguish patients infected with COVID-19 from other causes of pneumonia, and AI interpretation of pathology slides may detect specific genetic mutations not previously identified without additional molecular tests.^{11,14,23,24,30-32}

The second area in which AI can impact health care is improving workflow and efficiency. AI has improved surgery scheduling, saving significant revenue, and decreased patient wait times for appointments.¹ AI can screen and triage radiographs, allowing attention to be directed to critical patients. This use would be valuable in many busy clinical settings, such as the recent COVID-19 pandemic.^8,23 Similarly, AI can screen retina images to prioritize urgent conditions.²⁵ AI has improved pathologists’ efficiency when used to detect breast metastases.³³ Finally, AI may reduce medical errors, thereby ensuring patient safety.^8,9,34

A third health care benefit of AI is in public health and epidemiology. AI can assist with clinical decision-making and diagnoses in low-income countries and areas with limited health care resources and personnel.^25,29 AI can improve identification of infectious outbreaks, such as tuberculosis, malaria, dengue fever, and influenza.^29,35-40 AI has been used to predict transmission patterns of the Zika virus and the current COVID-19 pandemic.^41,42 Applications can stratify the risk of outbreaks based on multiple factors, including age, income, race, atypical geographic clusters, and seasonal factors like rainfall and temperature.^35,36,38,43 AI has been used to assess morbidity and mortality, such as predicting disease severity with malaria and identifying treatment failures in tuberculosis.²⁹

Finally, AI can dramatically impact health care due to processing large data sets or disconnected volumes of patient information—so-called big data.^44-46 An example is the widespread use of electronic health records (EHRs) such as the Computerized Patient Record System used in Veteran Affairs medical centers (VAMCs). Much of patient information exists in written text: HCP notes, laboratory and radiology reports, medication records, etc. Natural language processing (NLP) allows platforms to sort through extensive volumes of data on complex patients at rates much faster than human capability, which has great potential to assist with diagnosis and treatment decisions.⁹

Medical literature is being produced at rates that exceed our ability to digest. More than 200,000 cancer-related articles were published in 2019 alone.¹⁴ NLP capabilities of AI have the potential to rapidly sort through this extensive medical literature and relate specific verbiage in patient records guiding therapy.⁴⁶ IBM Watson, a supercomputer based on ML and NLP, demonstrates this concept with many potential applications, only some of which relate to health care.^1,9 Watson has an oncology component to assimilate multiple aspects of patient care, including clinical notes, pathology results, radiograph findings, staging, and a tumor’s genetic profile. It coordinates these inputs from the EHR and mines medical literature and research databases to recommend treatment options.^1,46 AI can assess and compile far greater patient data and therapeutic options than would be feasible by individual clinicians, thus providing customized patient care.⁴⁷ Watson has partnered with numerous medical centers, including MD Anderson Cancer Center and Memorial Sloan Kettering Cancer Center, with variable success.^44,47-49 While the full potential of Watson appears not yet realized, these AI-driven approaches will likely play an important role in leveraging the hidden value in the expanding volume of health care information.

Medical Specialty Applications

Radiology

Currently > 70% of FDA-approved AI medical devices are in the field of radiology.² Most radiology departments have used AI-friendly digital imaging for years, such as the picture archiving and communication systems used by numerous health care systems, including VAMCs.^2,15 Gray-scale images common in radiology lend themselves to standardization, although AI is not limited to black-and- white image interpretation.¹⁵

An abundance of literature describes plain radiograph interpretation using AI. One FDA-approved platform improved X-ray diagnosis of wrist fractures when used by emergency medicine clinicians.^2,50 AI has been applied to chest X-ray (CXR) interpretation of many conditions, including pneumonia, tuberculosis, malignant lung lesions, and COVID-19.^{23,25,28,44,51-53} For example, Nam and colleagues suggested AI is better at diagnosing malignant pulmonary nodules from CXRs than are trained radiologists.²⁸

In addition to plain radiographs, AI has been applied to many other imaging technologies, including ultrasounds, positron emission tomography, mammograms, computed tomography (CT), and magnetic resonance imaging (MRI).^{15,26,44,48,54-56} A large study demonstrated that ML platforms significantly reduced the time to diagnose intracranial hemorrhages on CT and identified subtle hemorrhages missed by radiologists.⁵⁵ Other studies have claimed that AI programs may be better than radiologists in detecting cancer in screening mammograms, and 3 FDA-approved devices focus on mammogram interpretation.^2,15,54,57 There is also great interest in MRI applications to detect and predict prognosis for breast cancer based on imaging findings.^21,56

Aside from providing accurate diagnoses, other studies focus on AI radiograph interpretation to assist with patient screening, triage, improving time to final diagnosis, providing a rapid “second opinion,” and even monitoring disease progression and offering insights into prognosis.^{8,21,23,52,55,56,58} These features help in busy urban centers but may play an even greater role in areas with limited access to health care or trained specialists such as radiologists.⁵²

Cardiology

Cardiology has the second highest number of FDA-approved AI applications.² Many cardiology AI platforms involve image analysis, as described in several recent reviews.^45,59,60 AI has been applied to echocardiography to measure ejection fractions, detect valvular disease, and assess heart failure from hypertrophic and restrictive cardiomyopathy and amyloidosis.^45,48,59 Applications for cardiac CT scans and CT angiography have successfully quantified both calcified and noncalcified coronary artery plaques and lumen assessments, assessed myocardial perfusion, and performed coronary artery calcium scoring.^45,59,60 Likewise, AI applications for cardiac MRI have been used to quantitate ejection fraction, large vessel flow assessment, and cardiac scar burden.^45,59

For years ECG devices have provided interpretation with limited accuracy using preprogrammed parameters.⁴⁸ However, the application of AI allows ECG interpretation on par with trained cardiologists. Numerous such AI applications exist, and 2 FDA-approved devices perform ECG interpretation.^2,61-64 One of these devices incorporates an AI-powered stethoscope to detect atrial fibrillation and heart murmurs.⁶⁵

Pathology

The advancement of whole slide imaging, wherein entire slides can be scanned and digitized at high speed and resolution, creates great potential for AI applications in pathology.^{12,24,32,33,66} A landmark study demonstrating the potential of AI for assessing whole slide imaging examined sentinel lymph node metastases in patients with breast cancer.²² Multiple algorithms in the study demonstrated that AI was equivalent or better than pathologists in detecting metastases, especially when the pathologists were time-constrained consistent with a normal working environment. Significantly, the most accurate and efficient diagnoses were achieved when the pathologist and AI interpretations were used together.^22,33

AI has shown promise in diagnosing many other entities, including cancers of the prostate (including Gleason scoring), lung, colon, breast, and skin.^{11,12,24,27,32,67} In addition, AI has shown great potential in scoring biomarkers important for prognosis and treatment, such as immunohistochemistry (IHC) labeling of Ki-67 and PD-L1.³² Pathologists can have difficulty classifying certain tumors or determining the site of origin for metastases, often having to rely on IHC with limited success. The unique features of image analysis with AI have the potential to assist in classifying difficult tumors and identifying sites of origin for metastatic disease based on morphology alone.¹¹

Oncology depends heavily on molecular pathology testing to dictate treatment options and determine prognosis. Preliminary studies suggest that AI interpretation alone has the potential to delineate whether certain molecular mutations are present in tumors from various sites.^11,14,24,32 One study combined histology and genomic results for AI interpretation that improved prognostic predictions.⁶⁸ In addition, AI analysis may have potential in predicting tumor recurrence or prognosis based on cellular features, as demonstrated for lung cancer and melanoma.^67,69,70

Ophthalmology

AI applications for ophthalmology have focused on diabetic retinopathy, age-related macular degeneration, glaucoma, retinopathy of prematurity, age-related and congenital cataracts, and retinal vein occlusion.^71-73 Diabetic retinopathy is a leading cause of blindness and has been studied by numerous platforms with good success, most having used color fundus photography.^71,72 One study showed AI could diagnose diabetic retinopathy and diabetic macular edema with specificities similar to ophthalmologists.⁷⁴ In 2018, the FDA approved the AI platform IDx-DR. This diagnostic system classifies retinal images and recommends referral for patients determined to have “more than mild diabetic retinopathy” and reexamination within a year for other patients.^8,75 Significantly, the platform recommendations do not require confirmation by a clinician.⁸

AI has been applied to other modalities in ophthalmology such as optical coherence tomography (OCT) to diagnose retinal disease and to predict appropriate management of congenital cataracts.^25,73,76 For example, an AI application using OCT has been demonstrated to match or exceed the accuracy of retinal experts in diagnosing and triaging patients with a variety of retinal pathologies, including patients needing urgent referrals.⁷⁷

Dermatology

Multiple studies demonstrate AI performs at least equal to experienced dermatologists in differentiating selected skin lesions.^78-81 For example, Esteva and colleagues demonstrated AI could differentiate keratinocyte carcinomas from benign seborrheic keratoses and malignant melanomas from benign nevi with accuracy equal to 21 board-certified dermatologists.⁷⁸

AI is applicable to various imaging procedures common to dermatology, such as dermoscopy, very high-frequency ultrasound, and reflectance confocal microscopy.⁸² Several studies have demonstrated that AI interpretation compared favorably to dermatologists evaluating dermoscopy to assess melanocytic lesions.^78-81,83

A limitation in these studies is that they differentiate only a few diagnoses.⁸² Furthermore, dermatologists have sensory input such as touch and visual examination under various conditions, something AI has yet to replicate.^15,34,84 Also, most AI devices use no or limited clinical information.⁸¹ Dermatologists can recognize rarer conditions for which AI models may have had limited or no training.³⁴ Nevertheless, a recent study assessed AI for the diagnosis of 134 separate skin disorders with promising results, including providing diagnoses with accuracy comparable to that of dermatologists and providing accurate treatment strategies.⁸⁴ As Topol points out, most skin lesions are diagnosed in the primary care setting where AI can have a greater impact when used in conjunction with the clinical impression, especially where specialists are in limited supply.^48,78

Finally, dermatology lends itself to using portable or smartphone applications (apps) wherein the user can photograph a lesion for analysis by AI algorithms to assess the need for further evaluation or make treatment recommendations.^34,84,85 Although results from currently available apps are not encouraging, they may play a greater role as the technology advances.^34,85

Oncology

Applications of AI in oncology include predicting prognosis for patients with cancer based on histologic and/or genetic information.^14,68,86 Programs can predict the risk of complications before and recurrence risks after surgery for malignancies.^44,87-89 AI can also assist in treatment planning and predict treatment failure with radiation therapy.^90,91

AI has great potential in processing the large volumes of patient data in cancer genomics. Next-generation sequencing has allowed for the identification of millions of DNA sequences in a single tumor to detect genetic anomalies.⁹² Thousands of mutations can be found in individual tumor samples, and processing this information and determining its significance can be beyond human capability.¹⁴ We know little about the effects of various mutation combinations, and most tumors have a heterogeneous molecular profile among different cell populations.^14,93 The presence or absence of various mutations can have diagnostic, prognostic, and therapeutic implications.⁹³ AI has great potential to sort through these complex data and identify actionable findings.

More than 200,000 cancer-related articles were published in 2019, and publications in the field of cancer genomics are increasing exponentially.^14,92,93 Patel and colleagues assessed the utility of IBM Watson for Genomics against results from a molecular tumor board.⁹³ Watson for Genomics identified potentially significant mutations not identified by the tumor board in 32% of patients. Most mutations were related to new clinical trials not yet added to the tumor board watch list, demonstrating the role AI will have in processing the large volume of genetic data required to deliver personalized medicine moving forward.

Gastroenterology

AI has shown promise in predicting risk or outcomes based on clinical parameters in various common gastroenterology problems, including gastric reflux, acute pancreatitis, gastrointestinal bleeding, celiac disease, and inflammatory bowel disease.^94,95 AI endoscopic analysis has demonstrated potential in assessing Barrett’s esophagus, gastric Helicobacter pylori infections, gastric atrophy, and gastric intestinal metaplasia.⁹⁵ Applications have been used to assess esophageal, gastric, and colonic malignancies, including depth of invasion based on endoscopic images.⁹⁵ Finally, studies have evaluated AI to assess small colon polyps during colonoscopy, including differentiating benign and premalignant polyps with success comparable to gastroenterologists.^94,95 AI has been shown to increase the speed and accuracy of gastroenterologists in detecting small polyps during colonoscopy.⁴⁸ In a prospective randomized study, colonoscopies performed using an AI device identified significantly more small adenomatous polyps than colonoscopies without AI.⁹⁶

Neurology

It has been suggested that AI technologies are well suited for application in neurology due to the subtle presentation of many neurologic diseases.¹⁶ Viz LVO, the first CMS-approved AI reimbursement for the diagnosis of strokes, analyzes CTs to detect early ischemic strokes and alerts the medical team, thus shortening time to treatment.^3,97 Many other AI platforms are in use or development that use CT and MRI for the early detection of strokes as well as for treatment and prognosis.^9,97

AI technologies have been applied to neurodegenerative diseases, such as Alzheimer and Parkinson diseases.^16,98 For example, several studies have evaluated patient movements in Parkinson disease for both early diagnosis and to assess response to treatment.⁹⁸ These evaluations included assessment with both external cameras as well as wearable devices and smartphone apps.

Mental Health

Due to the interactive nature of mental health care, the field has been slower to develop AI applications.¹⁸ With heavy reliance on textual information (eg, clinic notes, mood rating scales, and documentation of conversations), successful AI applications in this field will likely rely heavily on NLP.¹⁸ However, studies investigating the application of AI to mental health have also incorporated data such as brain imaging, smartphone monitoring, and social media platforms, such as Facebook and Twitter.^18,103,104

The risk of suicide is higher in veteran patients, and ML algorithms have had limited success in predicting suicide risk in both veteran and nonveteran populations.^104-106 While early models have low positive predictive values and low sensitivities, they still promise to be a useful tool in conjunction with traditional risk assessments.¹⁰⁶ Kessler and colleagues suggest that combining multiple rather than single ML algorithms might lead to greater success.^105,106

AI may assist in diagnosing other mental health disorders, including major depressive disorder, attention deficit hyperactivity disorder (ADHD), schizophrenia, posttraumatic stress disorder, and Alzheimer disease.^103,104,107 These investigations are in the early stages with limited clinical applicability. However, 2 AI applications awaiting FDA approval relate to ADHD and opioid use.² Furthermore, potential exists for AI to not only assist with prevention and diagnosis of ADHD, but also to identify optimal treatment options.^2,103

General and Personalized Medicine

Additional AI applications include diagnosing patients with suspected sepsis, measuring liver iron concentrations, predicting hospital mortality at the time of admission, and more.^2,108,109 AI can guide end-of-life decisions such as resuscitation status or whether to initiate mechanical ventilation.⁴⁸

AI-driven smartphone apps can be beneficial to both patients and clinicians. Examples include predicting nonadherence to anticoagulation therapy, monitoring heart rhythms for atrial fibrillation or signs of hyperkalemia in patients with renal failure, and improving outcomes for patients with diabetes mellitus by decreasing glycemic variability and reducing hypoglycemia.^8,48,110,111 The potential for AI applications to health care and personalized medicine are almost limitless.

Discussion

With ever-increasing expectations for all health care sectors to deliver timely, fiscally-responsible, high-quality health care, AI has the potential to have numerous impacts. AI can improve diagnostic accuracy while limiting errors and impact patient safety such as assisting with prescription delivery.^8,9,34 It can screen and triage patients, alerting clinicians to those needing more urgent evaluation.^8,23,77,97 AI also may increase a clinician’s efficiency and speed to render a diagnosis.^12,13,55,97 AI can provide a rapid second opinion, an ability especially beneficial in underserved areas with shortages of specialists.^{23,25,26,29,34} Similarly, AI may decrease the inter- and intraobserver variability common in many medical specialties.^12,27,45 AI applications can also monitor disease progression, identifying patients at greatest risk, and provide information for prognosis.^21,23,56,58 Finally, as described with applications using IBM Watson, AI can allow for an integrated approach to health care that is currently lacking.

We have described many reports suggesting AI can render diagnoses as well as or better than experienced clinicians, and speculation exists that AI will replace many roles currently performed by health care practitioners.^9,26 However, most studies demonstrate that AI’s diagnostic benefits are best realized when used to supplement a clinician’s impression.^{8,22,30,33,52,54,56,69,84} AI is not likely to replace humans in health care in the foreseeable future. The technology can be likened to the impact of CT scans developed in the 1970s in neurology. Prior to such detailed imaging, neurologists spent extensive time performing detailed physicals to render diagnoses and locate lesions before surgery. There was mistrust of this new technology and concern that CT scans would eliminate the need for neurologists.¹¹² On the contrary, neurology is alive and well, frequently being augmented by the technologies once speculated to replace it.

Commercial AI health care platforms represented a $2 billion industry in 2018 and are growing rapidly each year.^13,32 Many AI products are offered ready for implementation for various tasks, including diagnostics, patient management, and improved efficiency. Others will likely be provided as templates suitable for modification to meet the specific needs of the facility, practice, or specialty for its patient population.

AI Risks and Limitations

AI has several risks and limitations. Although there is progress in explainable AI, at times we still struggle to understand how the output provided by machine learning algorithms was created.^44,48 The many layers associated with deep learning self-determine the criteria to reach its conclusion, and these criteria can continually evolve. The parameters of deep learning are not preprogrammed, and there are too many individual data points to be extrapolated or deconvoluted for evaluation at our current level of knowledge.^26,51 These apparent lack of constraints cause concern for patient safety and suggest that greater validation and continued scrutiny of validity is required.^8,48 Efforts are underway to create explainable AI programs to make their processes more transparent, but such clarification is limited presently.^14,26,48,77

Another challenge of AI is determining the amount of training data required to function optimally. Also, if the output describes multiple variables or diagnoses, are each equally valid?¹¹³ Furthermore, many AI applications look for a specific process, such as cancer diagnoses on CXRs. However, how coexisting conditions like cardiomegaly, emphysema, pneumonia, etc, seen on CXRs will affect the diagnosis needs to be considered.^51,52 Zech and colleagues provide the example that diagnoses for pneumothorax are frequently rendered on CXRs with chest tubes in place.⁵¹ They suggest that CNNs may develop a bias toward diagnosing pneumothorax when chest tubes are present. Many current studies approach an issue in isolation, a situation not realistic in real-world clinical practice.²⁶

Most studies on AI have been retrospective, and frequently data used to train the program are preselected.^13,26 The data are typically validated on available databases rather than actual patients in the clinical setting, limiting confidence in the validity of the AI output when applied to real-world situations. Currently, fewer than 12 prospective trials had been published comparing AI with traditional clinical care.^13,114 Randomized prospective clinical trials are even fewer, with none currently reported from the United States.^13,114 The results from several studies have been shown to diminish when repeated prospectively.¹¹⁴

The FDA has created a new category known as Software as a Medical Device and has a Digital Health Innovation Action Plan to regulate AI platforms. Still, the process of AI regulation is of necessity different from traditional approval processes and is continually evolving.⁸ The FDA approval process cannot account for the fact that the program’s parameters may continually evolve or adapt.²

Guidelines for investigating and reporting AI research with its unique attributes are being developed. Examples include the TRIPOD-ML statement and others.^49,115 In September 2020, 2 publications addressed the paucity of gold-standard randomized clinical trials in clinical AI applications.^116,117 The SPIRIT-AI statement expands on the original SPIRIT statement published in 2013 to guide minimal reporting standards for AI clinical trial protocols to promote transparency of design and methodology.¹¹⁶ Similarly, the CONSORT-AI extension, stemming from the original CONSORT statement in 1996, aims to ensure quality reporting of randomized controlled trials in AI.¹¹⁷

Another risk with AI is that while an individual physician making a mistake may adversely affect 1 patient, a single mistake in an AI algorithm could potentially affect thousands of patients.⁴⁸ Also, AI programs developed for patient populations at a facility may not translate to another. Referred to as overfitting, this phenomenon relates to selection bias in training data sets.^{15,34,49,51,52} Studies have shown that programs that underrepresent certain group characteristics such as age, sex, or race may be less effective when applied to a population in which these characteristics have differing representations.^8,48,49 This problem of underrepresentation has been demonstrated in programs interpreting pathology slides, radiographs, and skin lesions.^15,32,51

Admittedly, most of these challenges are not specific to AI and existed in health care previously. Physicians make mistakes, treatments are sometimes used without adequate prospective studies, and medications are given without understanding their mechanism of action, much like AI-facilitated processes reach a conclusion that cannot be fully explained.⁴⁸

Conclusions

The view that AI will dramatically impact health care in the coming years will likely prove true. However, much work is needed, especially because of the paucity of prospective clinical trials as has been historically required in medical research. Any concern that AI will replace HCPs seems unwarranted. Early studies suggest that even AI programs that appear to exceed human interpretation perform best when working in cooperation with and oversight from clinicians. AI’s greatest potential appears to be its ability to augment care from health professionals, improving efficiency and accuracy, and should be anticipated with enthusiasm as the field moves forward at an exponential rate.

Acknowledgments

The authors thank Makenna G. Thomas for proofreading and review of the manuscript. This material is the result of work supported with resources and the use of facilities at the James A. Haley Veterans’ Hospital. This research has been approved by the James A. Haley Veteran’s Hospital Office of Communications and Media.

When I was a new palliative care nurse practitioner (NP), I remember my mentor telling me that communication in our field is equivalent to surgical procedures in general surgery. Conversations need to be handled with accuracy and precision, conducted in a timely fashion, and require skills that take practice to sharpen. Over the years, I learned that unlike surgery, we do not have control over how the procedure will flow. We approach patients with a blank canvas, open to receive messages that will be shared and reacted to accordingly. The ability to communicate effectively also requires compassion, which is a trait that tends to be inherent in humans and typically is not learned from textbooks but can be cultivated with training and application.

Among the barriers identified to effective communication are avoiding emotional issues and focusing on technical topics due in part to the fear of lengthy encounters, not allowing patients or families enough time to speak, and reframing instead of validating emotions.¹ Many years later, I had the chance to help care for a patient whose story reminds me of how our choice of words and how our interpretation of what we are told can influence the way we care for patients and their families.

Case Presentation

Mr. P, aged 86 years, was admitted to a teaching hospital for pneumonia and heart failure exacerbation. He was treated with diuretics and antibiotics and discharged home on room air after 3 days. He returned to the hospital after 8 days, reporting labored breathing. He was found to be hypoxic, and a further workup revealed acute hypoxic respiratory failure that was likely from severe pulmonary hypertension and exacerbation of his heart failure. Left heart disease is a common cause of pulmonary hypertension, which can lead to right ventricular failure and increased mortality.²

After meeting with his pulmonologist and cardiologist, Mr. P elected for a do-not-resuscitate code status and declined to be intubated. He also refused further diagnostics and life-prolonging treatments for his conditions, including a stress test, cardiac catheterization, and a right heart catheterization. He required bilevel positive airway pressure (BPAP) support at bedtime, which he also declined. He agreed to the use of supplemental oxygen through a nasal cannula and always needed 5 liters of oxygen.

Palliative care was consulted to assist with goals of care discussion. This visit took place during the COVID-19 pandemic, but Mr. P had tested negative for the COVID virus, so the palliative care NP was able to meet with Mr. P in person. He shared his understanding of the serious nature of his condition and the likelihood of a limited life expectancy without further diagnostics and possible life-prolonging treatments. He said his goal was to go home and spend the remainder of his life with his wife. He had not been out of bed since his hospitalization except to transfer to a nearby chair with the help of his nurse due to exertional dyspnea and generalized weakness. Prior to his recent hospitalizations, he was independently ambulating and had no dyspnea when performing strenuous activities. Mr. P shared that his wife was aged in her 70s and was legally blind. He added that she did not require physical assistance, but he was unsure whether she could help him because they had not been in such a situation previously. They had a daughter who visited frequently and helped with driving them to doctors’ appointments and shopping. Mr. P shared that he wanted to go home. After explaining the option of home hospice, Mr. P decided he wanted to receive hospice services at home and asked palliative care NP to contact his daughter to let her know his wishes and to tell her more about how hospice can help with his care.

The palliative care NP met with Mr. P’s nurse and shared the outcome of her visit. His nurse asked the palliative care NP whether she was familiar with his daughter. The nurse added that she wanted the palliative care NP to know that Mr. P’s daughter was quite angry and upset with his doctors after being told about his prognosis. His doctors’ notes also indicated that Mr. P wanted them to contact his daughter regarding his condition and plans for discharge, concluding that he deferred to his daughter for medical decision making.

As Mr. P’s hospitalization took place during the COVID pandemic, a face-to-face meeting with his family was not possible. The NP spoke with Mr. P’s daughter over the phone to relay his wishes and goals for his care. Mr. P’s daughter cried at times during the conversation and asked whether his condition was really that serious. The NP allowed Mr. P’s daughter to express her sadness and allowed for periods of silence during the conversation while his daughter gathered her composure. The NP reinforced the clinical information she had been provided by the medical team. Mr. P’s daughter added that he was completely independent, not requiring supplemental oxygen and was otherwise healthy just a month prior. She also asked whether there was truly nothing else that could be done to prolong his life. The NP acknowledged her observations and explained how Mr. P’s body and organs had not been able to bounce back from the recent insults to his overall physical condition.

After being told that Mr. P’s options for treatment were limited not only by his advanced age and comorbidities, but also the limitations and goals for his care he had identified, his daughter supported her father’s decision. The palliative care NP provided her information on how home hospice assists in her father’s care at home, including symptom management, nursing visits, home equipment, family support, among others. Mr. P’s daughter also said she would relay the information to her mother and call the palliative care NP if they had additional questions or concerns.

The outcome of her visit with Mr. P and his daughter were relayed by the palliative care NP to his acute health care team through an official response to the consultation request via his electronic health record. The palliative care NP also alerted the palliative care social worker to follow-up with Mr. P, his daughter, and his acute health care team to coordinate hospice services at the time of his discharge from the hospital.

Mr. P was discharged from the hospital with home hospice services after a few days. Three weeks later, Mr. P passed away peacefully on the in-patient unit of his home hospice agency as his physical care needs became too much for his family to provide at home a few days before his death. The palliative care social worker later shared with the NP that Mr. P’s daughter shared her gratitude and satisfaction with the care he had received not only from palliative care, but also from everyone during his hospitalization.

Discussion

Key themes found in end-of-life (EOL) communication with families and caregivers include highlighting clinical deterioration, involvement in decision making, continuation of high-quality care after cessation of aggressive measures, tailoring to individuals, clarity, honesty, and use of techniques in delivery.² Some of the techniques identified were pacing, staging, and repetition.³ Other techniques that can be beneficial include allowing for time to express one’s feelings, being comfortable with brief periods of silence, validating observations shared, among others. These themes were evident in the interactions that his health care team had with Mr. P and his daughter. With honesty and clarity, various members of the health care team repeatedly shared information regarding his clinical deterioration.

Family Influence

EOL decision-making roles within a family tend to originate from family interactional histories, familial roles as well as decision-making situations the family faces.⁴ The US medical and legal systems also recognize formal role assignments for surrogate decision makers.⁴ In the case of Mr. P, his advance directive (AD) identified his daughter as his surrogate decision maker. ADs are written statements made in advance by patients expressing their wishes and limitations for treatment as well as appointing surrogate decision makers when they become unable to decide for themselves in the future.⁵

During discussions about the goals for his care, Mr. P made his own medical decisions and elected to pursue a comfort-focused approach to care. His request for his health care team to reach out to his daughter was largely due to his need for assistance in explaining the complexity of his clinical condition to her and how hospice services would be helpful with his EOL care. Mr. P depended on his daughter to bring him to the hospital or to his doctors’ appointments, and she had been a major source of support for him and his wife. Contrary to the belief of some of his health care practitioners, Mr. P was not deferring his medical decisions to his daughter but rather allowing for her participation as his health care partner.

Communication between nurses and patients has been found to be challenging to both parties. Nurses express difficulties in areas that include supporting patients and families after they have had a difficult conversation with their physicians and responding to patients and family members’ emotions like anger.⁶ EOL care issues, such as family barriers to prognostic understanding, can interfere with psychosocial care.⁶ Families of patients approaching the EOL describe feeling mentally worn down and being unable to think straight, leading to feelings of being overwhelmed.⁷ They feel the need to be in a place where they can accept the content of difficult EOL conversations to be able to effectively engage.⁷

Studies have shown that family members of patients at the EOL experience stress, anxiety, fatigue and depression.⁸ Reactions that can be perceived as anger may not be so nor directed to the health care team. Questions raised regarding the accuracy of prognostication and treatment recommendations may not necessarily reflect concerns about the quality of care received but an exercise of advocacy in exploring other options on behalf of the patient. Allowing time for families to process the information received and react freely are necessary steps to facilitate reaching a place where they can acknowledge the information being relayed.

Communication Skills Training

Every member of the health care team should be equipped with the basic skills to have these conversations. The academic curricula for members of the health care team focuses on developing communication skills, but there has been a lack of content on palliative and EOL care.⁹

Due to time constraints and limited opportunities in the clinical setting, there has been an increasing use of simulation-based learning activities (SBLA) to enhance communication skills among nursing students.⁹ At this time, the impact of SBLA in enhancing communication competency is not fully known, but given the lack of clinical opportunities for students, this option is worth considering.⁹ When asked, nurses recognized the need for improved EOL communication education, training, and guidelines.¹⁰ They also felt that a multidisciplinary approach in EOL communication is beneficial. The inclusion of the End-of-Life Nursing Education Consortium (ELNEC) Core training in Bachelor of Science in Nursing programs have led to improved insight on palliative care and nurses’ role in palliative care and hospice among nursing students.¹¹

The Palliative Care and Hospice Education and Training Act of 2017 amended the Public Health Service Act to include improving EOL training for health care providers, including talking about death and dying.¹² Even though the Liaison Committee of Medical Education asked medical schools to incorporate EOL care education in the medical school curricula, there is still a lack of developmentally appropriate and supervised EOL education in medical schools.¹² Training on grief also has been lacking and less likely to be mandatory among medical students and residents: Workshops are mostly conducted before they can be applied in the clinical setting.¹³ Meanwhile, resources are available to assist physicians in EOL conversations with patient and families, such as the Serious Illness Conversation Guide, The Conversation Project, and Stanford’s Letter Project.¹²

Conclusions

Palliative consultation is associated with an overall improvement in EOL care, communication, and support, according to families of deceased patients.¹⁴ It has also been shown to enhance patients’ quality of life and mood, improve documentation of resuscitation preferences, and lead to less aggressive care at the EOL, including less chemotherapy.¹⁵ Integration of palliative care in the care of patients hospitalized with acute heart failure has been associated with improved quality of life, decreased symptom burden and depressive symptoms, and increased participation in advance care planning.¹⁶

The involvement of palliative care in the care of patients and their families at EOL enhances goals of care discussions that improve understanding for everyone involved. It helps provide consistency with the message being delivered by the rest of the health care team to patients and families regarding prognosis and recommendations. Palliative care can provide an alternative when all other aggressive measures are no longer helpful and allow for the continuation of care with a shift in focus from prolonging life to promoting its quality. Furthermore, palliative care involvement for care of patients with life-limiting illness also has been found to improve symptom control, decrease hospitalizations and health care costs, and even improve mortality.¹⁷A multidisciplinary approach to palliative care EOL conversations is beneficial, but every member of the health care team should have the training, education, and skills to be ready to have these difficult conversations. These health care team members include physicians, advance practice clinicians, nurses, social workers, and chaplains, among others. Patients and families are likely to be in contact with different members of the health care team who should be able to carry out therapeutic conversations. Using validated tools and resources on communication techniques through evidence-based practice is helpful and should be encouraged. This provides a framework on how EOL conversations should be conducted in the clinical setting to augment the identified lack of training on EOL communication in schools. Repeated opportunities for its use over time will help improve the ability of clinicians to engage in effective EOL communication.

When I was a new palliative care nurse practitioner (NP), I remember my mentor telling me that communication in our field is equivalent to surgical procedures in general surgery. Conversations need to be handled with accuracy and precision, conducted in a timely fashion, and require skills that take practice to sharpen. Over the years, I learned that unlike surgery, we do not have control over how the procedure will flow. We approach patients with a blank canvas, open to receive messages that will be shared and reacted to accordingly. The ability to communicate effectively also requires compassion, which is a trait that tends to be inherent in humans and typically is not learned from textbooks but can be cultivated with training and application.

Among the barriers identified to effective communication are avoiding emotional issues and focusing on technical topics due in part to the fear of lengthy encounters, not allowing patients or families enough time to speak, and reframing instead of validating emotions.¹ Many years later, I had the chance to help care for a patient whose story reminds me of how our choice of words and how our interpretation of what we are told can influence the way we care for patients and their families.

Case Presentation

Mr. P, aged 86 years, was admitted to a teaching hospital for pneumonia and heart failure exacerbation. He was treated with diuretics and antibiotics and discharged home on room air after 3 days. He returned to the hospital after 8 days, reporting labored breathing. He was found to be hypoxic, and a further workup revealed acute hypoxic respiratory failure that was likely from severe pulmonary hypertension and exacerbation of his heart failure. Left heart disease is a common cause of pulmonary hypertension, which can lead to right ventricular failure and increased mortality.²

After meeting with his pulmonologist and cardiologist, Mr. P elected for a do-not-resuscitate code status and declined to be intubated. He also refused further diagnostics and life-prolonging treatments for his conditions, including a stress test, cardiac catheterization, and a right heart catheterization. He required bilevel positive airway pressure (BPAP) support at bedtime, which he also declined. He agreed to the use of supplemental oxygen through a nasal cannula and always needed 5 liters of oxygen.

Palliative care was consulted to assist with goals of care discussion. This visit took place during the COVID-19 pandemic, but Mr. P had tested negative for the COVID virus, so the palliative care NP was able to meet with Mr. P in person. He shared his understanding of the serious nature of his condition and the likelihood of a limited life expectancy without further diagnostics and possible life-prolonging treatments. He said his goal was to go home and spend the remainder of his life with his wife. He had not been out of bed since his hospitalization except to transfer to a nearby chair with the help of his nurse due to exertional dyspnea and generalized weakness. Prior to his recent hospitalizations, he was independently ambulating and had no dyspnea when performing strenuous activities. Mr. P shared that his wife was aged in her 70s and was legally blind. He added that she did not require physical assistance, but he was unsure whether she could help him because they had not been in such a situation previously. They had a daughter who visited frequently and helped with driving them to doctors’ appointments and shopping. Mr. P shared that he wanted to go home. After explaining the option of home hospice, Mr. P decided he wanted to receive hospice services at home and asked palliative care NP to contact his daughter to let her know his wishes and to tell her more about how hospice can help with his care.

The palliative care NP met with Mr. P’s nurse and shared the outcome of her visit. His nurse asked the palliative care NP whether she was familiar with his daughter. The nurse added that she wanted the palliative care NP to know that Mr. P’s daughter was quite angry and upset with his doctors after being told about his prognosis. His doctors’ notes also indicated that Mr. P wanted them to contact his daughter regarding his condition and plans for discharge, concluding that he deferred to his daughter for medical decision making.

As Mr. P’s hospitalization took place during the COVID pandemic, a face-to-face meeting with his family was not possible. The NP spoke with Mr. P’s daughter over the phone to relay his wishes and goals for his care. Mr. P’s daughter cried at times during the conversation and asked whether his condition was really that serious. The NP allowed Mr. P’s daughter to express her sadness and allowed for periods of silence during the conversation while his daughter gathered her composure. The NP reinforced the clinical information she had been provided by the medical team. Mr. P’s daughter added that he was completely independent, not requiring supplemental oxygen and was otherwise healthy just a month prior. She also asked whether there was truly nothing else that could be done to prolong his life. The NP acknowledged her observations and explained how Mr. P’s body and organs had not been able to bounce back from the recent insults to his overall physical condition.

After being told that Mr. P’s options for treatment were limited not only by his advanced age and comorbidities, but also the limitations and goals for his care he had identified, his daughter supported her father’s decision. The palliative care NP provided her information on how home hospice assists in her father’s care at home, including symptom management, nursing visits, home equipment, family support, among others. Mr. P’s daughter also said she would relay the information to her mother and call the palliative care NP if they had additional questions or concerns.

The outcome of her visit with Mr. P and his daughter were relayed by the palliative care NP to his acute health care team through an official response to the consultation request via his electronic health record. The palliative care NP also alerted the palliative care social worker to follow-up with Mr. P, his daughter, and his acute health care team to coordinate hospice services at the time of his discharge from the hospital.

Mr. P was discharged from the hospital with home hospice services after a few days. Three weeks later, Mr. P passed away peacefully on the in-patient unit of his home hospice agency as his physical care needs became too much for his family to provide at home a few days before his death. The palliative care social worker later shared with the NP that Mr. P’s daughter shared her gratitude and satisfaction with the care he had received not only from palliative care, but also from everyone during his hospitalization.

Discussion

Key themes found in end-of-life (EOL) communication with families and caregivers include highlighting clinical deterioration, involvement in decision making, continuation of high-quality care after cessation of aggressive measures, tailoring to individuals, clarity, honesty, and use of techniques in delivery.² Some of the techniques identified were pacing, staging, and repetition.³ Other techniques that can be beneficial include allowing for time to express one’s feelings, being comfortable with brief periods of silence, validating observations shared, among others. These themes were evident in the interactions that his health care team had with Mr. P and his daughter. With honesty and clarity, various members of the health care team repeatedly shared information regarding his clinical deterioration.

Family Influence

EOL decision-making roles within a family tend to originate from family interactional histories, familial roles as well as decision-making situations the family faces.⁴ The US medical and legal systems also recognize formal role assignments for surrogate decision makers.⁴ In the case of Mr. P, his advance directive (AD) identified his daughter as his surrogate decision maker. ADs are written statements made in advance by patients expressing their wishes and limitations for treatment as well as appointing surrogate decision makers when they become unable to decide for themselves in the future.⁵

During discussions about the goals for his care, Mr. P made his own medical decisions and elected to pursue a comfort-focused approach to care. His request for his health care team to reach out to his daughter was largely due to his need for assistance in explaining the complexity of his clinical condition to her and how hospice services would be helpful with his EOL care. Mr. P depended on his daughter to bring him to the hospital or to his doctors’ appointments, and she had been a major source of support for him and his wife. Contrary to the belief of some of his health care practitioners, Mr. P was not deferring his medical decisions to his daughter but rather allowing for her participation as his health care partner.

Communication between nurses and patients has been found to be challenging to both parties. Nurses express difficulties in areas that include supporting patients and families after they have had a difficult conversation with their physicians and responding to patients and family members’ emotions like anger.⁶ EOL care issues, such as family barriers to prognostic understanding, can interfere with psychosocial care.⁶ Families of patients approaching the EOL describe feeling mentally worn down and being unable to think straight, leading to feelings of being overwhelmed.⁷ They feel the need to be in a place where they can accept the content of difficult EOL conversations to be able to effectively engage.⁷

Studies have shown that family members of patients at the EOL experience stress, anxiety, fatigue and depression.⁸ Reactions that can be perceived as anger may not be so nor directed to the health care team. Questions raised regarding the accuracy of prognostication and treatment recommendations may not necessarily reflect concerns about the quality of care received but an exercise of advocacy in exploring other options on behalf of the patient. Allowing time for families to process the information received and react freely are necessary steps to facilitate reaching a place where they can acknowledge the information being relayed.

Communication Skills Training

Every member of the health care team should be equipped with the basic skills to have these conversations. The academic curricula for members of the health care team focuses on developing communication skills, but there has been a lack of content on palliative and EOL care.⁹

Due to time constraints and limited opportunities in the clinical setting, there has been an increasing use of simulation-based learning activities (SBLA) to enhance communication skills among nursing students.⁹ At this time, the impact of SBLA in enhancing communication competency is not fully known, but given the lack of clinical opportunities for students, this option is worth considering.⁹ When asked, nurses recognized the need for improved EOL communication education, training, and guidelines.¹⁰ They also felt that a multidisciplinary approach in EOL communication is beneficial. The inclusion of the End-of-Life Nursing Education Consortium (ELNEC) Core training in Bachelor of Science in Nursing programs have led to improved insight on palliative care and nurses’ role in palliative care and hospice among nursing students.¹¹

The Palliative Care and Hospice Education and Training Act of 2017 amended the Public Health Service Act to include improving EOL training for health care providers, including talking about death and dying.¹² Even though the Liaison Committee of Medical Education asked medical schools to incorporate EOL care education in the medical school curricula, there is still a lack of developmentally appropriate and supervised EOL education in medical schools.¹² Training on grief also has been lacking and less likely to be mandatory among medical students and residents: Workshops are mostly conducted before they can be applied in the clinical setting.¹³ Meanwhile, resources are available to assist physicians in EOL conversations with patient and families, such as the Serious Illness Conversation Guide, The Conversation Project, and Stanford’s Letter Project.¹²

Conclusions

Palliative consultation is associated with an overall improvement in EOL care, communication, and support, according to families of deceased patients.¹⁴ It has also been shown to enhance patients’ quality of life and mood, improve documentation of resuscitation preferences, and lead to less aggressive care at the EOL, including less chemotherapy.¹⁵ Integration of palliative care in the care of patients hospitalized with acute heart failure has been associated with improved quality of life, decreased symptom burden and depressive symptoms, and increased participation in advance care planning.¹⁶

The involvement of palliative care in the care of patients and their families at EOL enhances goals of care discussions that improve understanding for everyone involved. It helps provide consistency with the message being delivered by the rest of the health care team to patients and families regarding prognosis and recommendations. Palliative care can provide an alternative when all other aggressive measures are no longer helpful and allow for the continuation of care with a shift in focus from prolonging life to promoting its quality. Furthermore, palliative care involvement for care of patients with life-limiting illness also has been found to improve symptom control, decrease hospitalizations and health care costs, and even improve mortality.¹⁷A multidisciplinary approach to palliative care EOL conversations is beneficial, but every member of the health care team should have the training, education, and skills to be ready to have these difficult conversations. These health care team members include physicians, advance practice clinicians, nurses, social workers, and chaplains, among others. Patients and families are likely to be in contact with different members of the health care team who should be able to carry out therapeutic conversations. Using validated tools and resources on communication techniques through evidence-based practice is helpful and should be encouraged. This provides a framework on how EOL conversations should be conducted in the clinical setting to augment the identified lack of training on EOL communication in schools. Repeated opportunities for its use over time will help improve the ability of clinicians to engage in effective EOL communication.

When I was a new palliative care nurse practitioner (NP), I remember my mentor telling me that communication in our field is equivalent to surgical procedures in general surgery. Conversations need to be handled with accuracy and precision, conducted in a timely fashion, and require skills that take practice to sharpen. Over the years, I learned that unlike surgery, we do not have control over how the procedure will flow. We approach patients with a blank canvas, open to receive messages that will be shared and reacted to accordingly. The ability to communicate effectively also requires compassion, which is a trait that tends to be inherent in humans and typically is not learned from textbooks but can be cultivated with training and application.

Among the barriers identified to effective communication are avoiding emotional issues and focusing on technical topics due in part to the fear of lengthy encounters, not allowing patients or families enough time to speak, and reframing instead of validating emotions.¹ Many years later, I had the chance to help care for a patient whose story reminds me of how our choice of words and how our interpretation of what we are told can influence the way we care for patients and their families.

Case Presentation

Mr. P, aged 86 years, was admitted to a teaching hospital for pneumonia and heart failure exacerbation. He was treated with diuretics and antibiotics and discharged home on room air after 3 days. He returned to the hospital after 8 days, reporting labored breathing. He was found to be hypoxic, and a further workup revealed acute hypoxic respiratory failure that was likely from severe pulmonary hypertension and exacerbation of his heart failure. Left heart disease is a common cause of pulmonary hypertension, which can lead to right ventricular failure and increased mortality.²

After meeting with his pulmonologist and cardiologist, Mr. P elected for a do-not-resuscitate code status and declined to be intubated. He also refused further diagnostics and life-prolonging treatments for his conditions, including a stress test, cardiac catheterization, and a right heart catheterization. He required bilevel positive airway pressure (BPAP) support at bedtime, which he also declined. He agreed to the use of supplemental oxygen through a nasal cannula and always needed 5 liters of oxygen.

Palliative care was consulted to assist with goals of care discussion. This visit took place during the COVID-19 pandemic, but Mr. P had tested negative for the COVID virus, so the palliative care NP was able to meet with Mr. P in person. He shared his understanding of the serious nature of his condition and the likelihood of a limited life expectancy without further diagnostics and possible life-prolonging treatments. He said his goal was to go home and spend the remainder of his life with his wife. He had not been out of bed since his hospitalization except to transfer to a nearby chair with the help of his nurse due to exertional dyspnea and generalized weakness. Prior to his recent hospitalizations, he was independently ambulating and had no dyspnea when performing strenuous activities. Mr. P shared that his wife was aged in her 70s and was legally blind. He added that she did not require physical assistance, but he was unsure whether she could help him because they had not been in such a situation previously. They had a daughter who visited frequently and helped with driving them to doctors’ appointments and shopping. Mr. P shared that he wanted to go home. After explaining the option of home hospice, Mr. P decided he wanted to receive hospice services at home and asked palliative care NP to contact his daughter to let her know his wishes and to tell her more about how hospice can help with his care.

The palliative care NP met with Mr. P’s nurse and shared the outcome of her visit. His nurse asked the palliative care NP whether she was familiar with his daughter. The nurse added that she wanted the palliative care NP to know that Mr. P’s daughter was quite angry and upset with his doctors after being told about his prognosis. His doctors’ notes also indicated that Mr. P wanted them to contact his daughter regarding his condition and plans for discharge, concluding that he deferred to his daughter for medical decision making.

As Mr. P’s hospitalization took place during the COVID pandemic, a face-to-face meeting with his family was not possible. The NP spoke with Mr. P’s daughter over the phone to relay his wishes and goals for his care. Mr. P’s daughter cried at times during the conversation and asked whether his condition was really that serious. The NP allowed Mr. P’s daughter to express her sadness and allowed for periods of silence during the conversation while his daughter gathered her composure. The NP reinforced the clinical information she had been provided by the medical team. Mr. P’s daughter added that he was completely independent, not requiring supplemental oxygen and was otherwise healthy just a month prior. She also asked whether there was truly nothing else that could be done to prolong his life. The NP acknowledged her observations and explained how Mr. P’s body and organs had not been able to bounce back from the recent insults to his overall physical condition.

After being told that Mr. P’s options for treatment were limited not only by his advanced age and comorbidities, but also the limitations and goals for his care he had identified, his daughter supported her father’s decision. The palliative care NP provided her information on how home hospice assists in her father’s care at home, including symptom management, nursing visits, home equipment, family support, among others. Mr. P’s daughter also said she would relay the information to her mother and call the palliative care NP if they had additional questions or concerns.

The outcome of her visit with Mr. P and his daughter were relayed by the palliative care NP to his acute health care team through an official response to the consultation request via his electronic health record. The palliative care NP also alerted the palliative care social worker to follow-up with Mr. P, his daughter, and his acute health care team to coordinate hospice services at the time of his discharge from the hospital.

Mr. P was discharged from the hospital with home hospice services after a few days. Three weeks later, Mr. P passed away peacefully on the in-patient unit of his home hospice agency as his physical care needs became too much for his family to provide at home a few days before his death. The palliative care social worker later shared with the NP that Mr. P’s daughter shared her gratitude and satisfaction with the care he had received not only from palliative care, but also from everyone during his hospitalization.

Discussion

Key themes found in end-of-life (EOL) communication with families and caregivers include highlighting clinical deterioration, involvement in decision making, continuation of high-quality care after cessation of aggressive measures, tailoring to individuals, clarity, honesty, and use of techniques in delivery.² Some of the techniques identified were pacing, staging, and repetition.³ Other techniques that can be beneficial include allowing for time to express one’s feelings, being comfortable with brief periods of silence, validating observations shared, among others. These themes were evident in the interactions that his health care team had with Mr. P and his daughter. With honesty and clarity, various members of the health care team repeatedly shared information regarding his clinical deterioration.

Family Influence

EOL decision-making roles within a family tend to originate from family interactional histories, familial roles as well as decision-making situations the family faces.⁴ The US medical and legal systems also recognize formal role assignments for surrogate decision makers.⁴ In the case of Mr. P, his advance directive (AD) identified his daughter as his surrogate decision maker. ADs are written statements made in advance by patients expressing their wishes and limitations for treatment as well as appointing surrogate decision makers when they become unable to decide for themselves in the future.⁵

During discussions about the goals for his care, Mr. P made his own medical decisions and elected to pursue a comfort-focused approach to care. His request for his health care team to reach out to his daughter was largely due to his need for assistance in explaining the complexity of his clinical condition to her and how hospice services would be helpful with his EOL care. Mr. P depended on his daughter to bring him to the hospital or to his doctors’ appointments, and she had been a major source of support for him and his wife. Contrary to the belief of some of his health care practitioners, Mr. P was not deferring his medical decisions to his daughter but rather allowing for her participation as his health care partner.

Communication between nurses and patients has been found to be challenging to both parties. Nurses express difficulties in areas that include supporting patients and families after they have had a difficult conversation with their physicians and responding to patients and family members’ emotions like anger.⁶ EOL care issues, such as family barriers to prognostic understanding, can interfere with psychosocial care.⁶ Families of patients approaching the EOL describe feeling mentally worn down and being unable to think straight, leading to feelings of being overwhelmed.⁷ They feel the need to be in a place where they can accept the content of difficult EOL conversations to be able to effectively engage.⁷

Studies have shown that family members of patients at the EOL experience stress, anxiety, fatigue and depression.⁸ Reactions that can be perceived as anger may not be so nor directed to the health care team. Questions raised regarding the accuracy of prognostication and treatment recommendations may not necessarily reflect concerns about the quality of care received but an exercise of advocacy in exploring other options on behalf of the patient. Allowing time for families to process the information received and react freely are necessary steps to facilitate reaching a place where they can acknowledge the information being relayed.

Communication Skills Training

Every member of the health care team should be equipped with the basic skills to have these conversations. The academic curricula for members of the health care team focuses on developing communication skills, but there has been a lack of content on palliative and EOL care.⁹

Due to time constraints and limited opportunities in the clinical setting, there has been an increasing use of simulation-based learning activities (SBLA) to enhance communication skills among nursing students.⁹ At this time, the impact of SBLA in enhancing communication competency is not fully known, but given the lack of clinical opportunities for students, this option is worth considering.⁹ When asked, nurses recognized the need for improved EOL communication education, training, and guidelines.¹⁰ They also felt that a multidisciplinary approach in EOL communication is beneficial. The inclusion of the End-of-Life Nursing Education Consortium (ELNEC) Core training in Bachelor of Science in Nursing programs have led to improved insight on palliative care and nurses’ role in palliative care and hospice among nursing students.¹¹

The Palliative Care and Hospice Education and Training Act of 2017 amended the Public Health Service Act to include improving EOL training for health care providers, including talking about death and dying.¹² Even though the Liaison Committee of Medical Education asked medical schools to incorporate EOL care education in the medical school curricula, there is still a lack of developmentally appropriate and supervised EOL education in medical schools.¹² Training on grief also has been lacking and less likely to be mandatory among medical students and residents: Workshops are mostly conducted before they can be applied in the clinical setting.¹³ Meanwhile, resources are available to assist physicians in EOL conversations with patient and families, such as the Serious Illness Conversation Guide, The Conversation Project, and Stanford’s Letter Project.¹²

Conclusions

Palliative consultation is associated with an overall improvement in EOL care, communication, and support, according to families of deceased patients.¹⁴ It has also been shown to enhance patients’ quality of life and mood, improve documentation of resuscitation preferences, and lead to less aggressive care at the EOL, including less chemotherapy.¹⁵ Integration of palliative care in the care of patients hospitalized with acute heart failure has been associated with improved quality of life, decreased symptom burden and depressive symptoms, and increased participation in advance care planning.¹⁶

The involvement of palliative care in the care of patients and their families at EOL enhances goals of care discussions that improve understanding for everyone involved. It helps provide consistency with the message being delivered by the rest of the health care team to patients and families regarding prognosis and recommendations. Palliative care can provide an alternative when all other aggressive measures are no longer helpful and allow for the continuation of care with a shift in focus from prolonging life to promoting its quality. Furthermore, palliative care involvement for care of patients with life-limiting illness also has been found to improve symptom control, decrease hospitalizations and health care costs, and even improve mortality.¹⁷A multidisciplinary approach to palliative care EOL conversations is beneficial, but every member of the health care team should have the training, education, and skills to be ready to have these difficult conversations. These health care team members include physicians, advance practice clinicians, nurses, social workers, and chaplains, among others. Patients and families are likely to be in contact with different members of the health care team who should be able to carry out therapeutic conversations. Using validated tools and resources on communication techniques through evidence-based practice is helpful and should be encouraged. This provides a framework on how EOL conversations should be conducted in the clinical setting to augment the identified lack of training on EOL communication in schools. Repeated opportunities for its use over time will help improve the ability of clinicians to engage in effective EOL communication.

Immune checkpoint inhibitors (CPIs) have revolutionized cancer therapy and improved the prognosis for a variety of advanced solid tumors and Hodgkin lymphoma, but evidence is growing regarding severe endocrine disturbances.^1,2 CPIs block inhibitory molecules on activated T cells to increase tumor cell destruction but also can breach normal tolerance, resulting in a spectrum of immune-related adverse events (irAE).^1,2 Programmed cell death-1 (PD-1) inhibitors are one type of CPIs. Pembrolizumab is a humanized monoclonal antibody that targets the PD-1 checkpoint pathway and is approved for the treatment of malignant melanoma and non-small cell lung cancer.^3,4 When the PD-1 checkpoint pathway is inhibited, T cells targeting cancer are activated, as are autoreactive T cells, such as those regulating pancreatic islet cell survival, which can lead to type 1 diabetes mellitus (T1DM).⁵

Case Presentation

A 95-year-old male veteran with long-standing, stable prediabetes was treated with pembrolizumab for stage 4 melanoma. Four months after treatment initiation and 3 weeks after completion of his sixth treatment cycle of pembrolizumab (2 mg/kg every 3 weeks), he presented for surveillance positron emission tomography (PET) and was incidentally found to have a serum glucose of 423 mg/dL. Hypothesis-driven history taking revealed polyuria, polydipsia, and a 12-lb weight loss during the previous 3 months. The patient reported no abdominal pain, nausea, or vomiting. He showed no evidence of pancreatic metastases on recent imaging. His family history was notable for a daughter with T1DM diagnosed at a young age.

On examination, the patient’s vital signs were normal aside from a blood pressure of 80/40 mm Hg. His body mass index was 30. He was alert and oriented with comfortable respirations and no Kussmaul breathing. He exhibited dry mucous membranes and poor skin turgor. Laboratory studies revealed 135 mmol/L sodium (reference, 135-145), 4.6 mmol/L potassium (reference, 3.6-5.2), 100 mmol/L chloride (reference, 99-106), bicarbonate of 26.5 mmol/L (reference, 23-29), serum blood urea nitrogen 27 mg/dL (reference, 6-24), 1.06 mg/dL creatinine (reference, 0.74-1.35), and 423 mg/dL glucose (reference, 70-100), with negative urine ketones. Further studies demonstrated 462 µmol/L fructosamine (reference, 190-270), correlating with hemoglobin A_1c (HbA_1c) close to 11.0% (HbA_1c was drawn on admission but cancelled by the laboratory for unknown reasons).^6,7 Later, an inappropriately low C-peptide level of 0.56 ng/mL (reference, 0.8-3.85) and a negative antiglutamic acid decarboxylase (GAD) antibody titer resulted. The patient was given IV hydration and admitted to the hospital. With input from endocrinology, the patient was started on 0.3 units per kg of body weight basal-prandial insulin therapy. Pembrolizumab was held. Six weeks after discharge, his HbA_1c was 7.2%, and C-peptide improved to 1.95 ng/mL and plasma glucose 116 mg/dL. After shared decision making with his health care team, the patient decided against restarting pembrolizumab. The patient reported that his functional status was preserved, and he preferred to take fewer medications at his advanced age. He died comfortably 6 months after this presentation from complications of metastatic melanoma.

Dicussion

Immunotherapy is now an integral part of cancer treatment and can result in endocrine disturbances.^1,2 Life-threatening irAEs are rare and may mimic more common conditions; thus, there is growing recognition of the need to educate health care professionals in appropriate screening and management of these conditions. CPI-induced T1DM is an uncommon but clinically significant event with an incidence of 0.4 to 1.27% and a median onset of 20 weeks after initiation of therapy (range, 1-228 weeks).^8-12In case seriesfrom 3 academic centers, 59 to81% of patients with CPI-induced T1DM presented with diabetic ketoacidosis (DKA), and only 40 to 71% of patients were autoantibody positive.^13-16 These patients are older than those presenting with classic T1DM, often require intensive care unit admission, and nearly invariably require exogenous insulin injections for metabolic control.^13-16

Based on the later age of onset of cancers that may be treated with CPI, patients with CPI-induced T1DM may be misdiagnosed with T2DM or hyperglycemia from other causes, such as medications or acute illness in the outpatient setting, risking suboptimal treatment.

Given the infrequent incidence and lack of controlled trials, screening and treatment recommendations for CPI-induced T1DM are based on principles derived from case series and expert opinion. Development of polyuria, polydipsia, weight loss, nausea, and/or vomiting should prompt investigation for possible development or worsening of hyperglycemia, suggestive of development of T1DM.¹⁷ American Society of Clinical Oncology (ASCO) guidelines recommend that serum glucose be assessed at baseline and with each treatment cycle during induction for 12 weeks, then every 3 to 6 weeks thereafter.¹⁷ There is no reported association between the number of CPI treatments and the development of DM.^8,9 Following our patient’s fifth pembrolizumab cycle, a random glucose reading was noted to be 186 mg/dL (Figure 1). Under the ASCO guidelines, ideally the patient would have received close clinical follow-up given the striking increase in plasma glucose compared with prior baseline lower values and perhaps been further evaluated with an anti-GAD antibody titer to screen for T1DM.¹⁷

Conclusions

Pembrolizumab-induced T1DM is a rare, but potentially life-threatening irAE. The acute risk of DKA requires early recognition and prompt treatment of patients taking CPIs. More than 90% of primary care physicians (PCPs) fulfill general medical care roles for patients with cancer; therefore, they play an essential role in evaluating symptoms during therapy.²⁰ Further studies evaluating the role of PCPs and outcomes when PCPs are involved in oncologic care should be conducted.

With increased index of suspicion, this clinical scenario presents an opportunity for PCPs that may help reduce irAE-associated morbidity and mortality of patients on CPIs, like pembrolizumab. Figure 2 illustrates an example addendum that can be used to alert and tag a PCP of a mutual patient after initiation of CPI therapy. Determining the optimal interface between PCPs, oncologists, and endocrinologists in delivering and coordinating high-quality cancer care in the setting of immunotherapy is an important area for ongoing quality improvement.

Acknowledgment

The authors thank all the staff and health care professionals at VA Greater Los Angeles Healthcare System who were involved in the care of this patient.