Clinical Review

The 2020 pandemic of coronavirus disease 2019 (COVID-19) has presented significant challenges to the health care workforce.^1,2 As New York City and its environs became the epicenter of the pandemic in the United States, we continued to care for our patients while simultaneously maintaining the education and well-being of our residents.³ Keeping this balance significantly strained resources and presented new challenges for education and service in residency education. What first emerged as an acute emergency has become a chronic disruption in the clinical learning environment. Programs are working to respond to the critical patient needs while ensuring continued progress toward training goals.

Since pregnancy is one condition for which healthy patients continued to require both outpatient visits and inpatient hospitalization, volume was not anticipated to be significantly decreased on our units. Thus, our ObGyn residency programs sought to expeditiously restructure our workforce and educational methods to address the demands of the pandemic. We were aided in our efforts by the Accreditation Council for Graduate Medical Education (ACGME) Extraordinary Circumstances policy. Our institutions were deemed to be functioning at Stage 3 Pandemic Emergency Status, a state in which “the increase in volume and/or severity of illness creates an extraordinary circumstance where routine care, education, and delivery must be reconfigured to focus only on patient care.”⁴

As of May 18, 2020, 26% of residency and fellowship programs in the United States were under Stage 3 COVID-19 Pandemic Emergency Status.⁵ Accordingly, our patient care delivery and educational processes were reconfigured within the context of Stage 3 Status, governed by the overriding principles of ensuring appropriate resources and training, adhering to work hour limits, providing adequate supervision, and credentialing fellows to function in our core specialty.

As ObGyn education leaders from 5 academic medical centers within the COVID-19 epicenter, we present a summary of best practices, based on our experiences, for each of the 4 categories of Stage 3 Status outlined by the ACGME. In an era of globalization, we must learn from pandemics, a call made after the Ebola outbreak in 2015.⁶ We recognize that this type of disruption could happen again with a possible second wave of COVID-19 or another emerging disease.⁷ Thus, we emphasize “lessons learned” that are applicable to a wide range of residency training programs facing various clinical crises.

Ensuring adequate resources and training

Within the context of Stage 3 Status, residency programs have the flexibility to increase residents’ availability in the clinical care setting. However, programs must ensure the safety of both patients and residents.

Continue to: Measures to decrease risk of infection...

One critical resource needed to protect patients and residents is personal protective equipment (PPE). Online instruction and in-person training were used to educate residents and staff on appropriate techniques for donning, doffing, and conserving PPE. Surgical teams were limited to 1 surgeon and 1 resident in each case. In an effort to limit direct contact with COVID-19 infected patients, the number of health care providers rounding on inpatients was restricted, and phone or video conversations were used for communication.

The workforce was modified to decrease exposure to infection and maintain a reserve of healthy residents who were working from home—anticipating that some residents would become ill and this reserve would be called for duty. Similar to other specialties, our programs organized the workforce by arranging residents into teams in which residents worked a number of shifts in a row.^8-12 Regular block schedules were disrupted and non-core rotations were deferred.

As surgeries were canceled and outpatient visits curtailed, many rotations required less resident coverage. Residents were reassigned from rotations where clinical work was suspended to accommodate increased staffing needs in other areas, while accounting for residents who were ill or on leave for postexposure quarantine. Typically, residents worked 12-hour shifts for 3 to 6 days followed by several days off or days working remotely. This team-based strategy decreased the number of residents exposed to COVID-19 at one time, provided time for recuperation, encouraged camaraderie, and enabled residents working remotely to coordinate care and participate in telehealth without direct patient contact.

To minimize high-risk exposure of pregnant residents or residents with underlying health conditions, these residents also worked remotely. Similar to other specialties, it was important to determine essential resident duties and enlist assistance from other clinicians, such as fellows, nurse practitioners, physician assistants, and midwives.

To protect residents and patients, maximizing testing of patients for COVID-19 was an important strategy. Based on early experience at 1 center with patients who were initially asymptomatic but later developed symptoms and tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), universal testing was implemented and endorsed by the New York State COVID-19 Maternity Task Force.¹³ Notably, 87.9% of patients who were positive for SARS-CoV-2 at the time of admission had no symptoms of COVID-19 at presentation. Because the asymptomatic carrier rate appears to be high in obstetric patients, testing of patients is paramount.^3,14 Finally, suspending visitation (except for 1 support person) also was instrumental in decreasing the risk of infection to residents.¹³

Resources for residents with COVID-19

This pandemic placed residency program directors in an unusual situation as frontline caregivers for their own residents. It was imperative to track residents with physical symptoms, conduct testing when possible, and follow the course of residents with confirmed or suspected COVID-19. As serious illness and death have been reported among otherwise healthy young people, we ensured that our homebound residents were frequently monitored.¹⁵ At several of our centers, residents with COVID-19 from any program who chose to separate from their families were provided with alternative housing accommodations. In addition, some of our graduate medical education offices identified specific physicians to care for residents with COVID-19 who did not require hospitalization.

Continue to: Deployment to other specialties...

Several hospitals in the United States redeployed residents because of staffing shortages in high-impact settings.¹² It was important for ObGyns to emphasize that the labor and delivery unit functions as the emergency ward for pregnant women, and that ObGyn residents possess skills specific to the care of these patients.

For our departments, we highlighted that external redeployment could adversely affect our workforce restructuring and, ultimately, patient care. We focused efforts on internal deployment or reassignment as much as possible. Some faculty and fellows in nonobstetric subspecialty areas were redirected to provide care on our inpatient obstetric services.

Educating residents

To maintain educational efforts with social distancing, we used videoconferencing to preserve the protected didactic education time that existed for our residents before the pandemic. This regularly scheduled, nonclinical time also was utilized to instruct residents on the rapidly changing clinical guidelines and to disseminate information about new institutional policies and procedures, ensuring that residents were adequately prepared for their new clinical work.

Work hour requirements

The ACGME requires that work hour limitations remain unchanged during Stage 3 Pandemic Emergency Status. As the pandemic presented new challenges and stressors for residents inside and outside the workplace, ensuring adequate time off to rest and recover was critical for maintaining the resident workforce’s health and wellness.

Thus, our workforce restructuring plans accounted for work hour limitations. As detailed above, the restructuring was accomplished by cohorting residents into small teams that remained unchanged for several weeks. Most shifts were limited to 12 hours, residents continued to be assigned at least 1 day off each week, and daily schedules were structured to ensure at least 10 hours off between shifts. Time spent working remotely was included in work hour calculations.

In addition, residents on “jeopardy” who were available for those who needed to be removed from direct patient care were given at least 1 day off per week in which they could not be pulled for clinical duty. Finally, prolonged inpatient assignments were limited; after these assignments, residents were given increased time for rest and recuperation.

Ensuring adequate supervision

The expectation during Stage 3 Pandemic Emergency Status is that residents, with adequate supervision, provide care that is appropriate for their level of training. To adequately and safely supervise residents, faculty needed training to remain well informed about the clinical care of COVID-19 patients. This was accomplished through frequent communication and consultation with colleagues in infectious disease, occupational health, and guidance from national organizations, such as the American College of Obstetricians and Gynecologists and the Centers for Disease Control and Prevention, and information from our state health departments.

Faculty members were trained in safe donning and doffing of PPE and infection control strategies to ensure they could safely oversee and train residents in these practices. Faculty schedules were significantly altered to ensure an adequate workforce and adequate resident supervision. Faculty efforts were focused on areas of critical need—in our case inpatient obstetrics—with a smaller workforce assigned to outpatient services and inpatient gynecology and gynecologic oncology. Many ObGyn subspecialist faculty were redeployed to general ObGyn inpatient units, thus permitting appropriate resident supervision at all times. In the outpatient setting, faculty adjusted to the changing demands and learned to conduct and supervise telehealth visits.

Finally, for those whose residents were deployed to other services (for example, internal medicine, emergency medicine, or critical care), supervision became paramount. We checked in with our deployed residents daily to be sure that their supervision on those services was adequate. Considering the extreme complexity, rapidly changing understanding of the disease, and often tragic patient outcomes, it was essential to ensure appropriate support and supervision on “off service” deployment.

Continue to: Fellows functioning in core specialty...

Fellows functioning in core specialty

Anticipating the increased need for clinicians on the obstetric services, fellows in subspecialty areas were granted emergency privileges to act as attending faculty in the core specialty, supervising residents and providing patient care. On the other hand, some of those fellows, primarily in gynecologic oncology, were externally redeployed out of core specialty to internal medicine and critical care units. Careful consideration of the fellows’ needs for supervision and support in these roles was essential, and similar support measures that were put in place for our residents were offered to fellows.

In conclusion

The COVID-19 pandemic has presented diverse and complex challenges to the entire health care workforce. Because this crisis is widespread and likely will be lengthy, a sustained and organized response is required.¹⁶ We have highlighted unique challenges specific to residency programs and presented collective best practices from our experiences in ObGyn navigating these obstacles, which are applicable to many other programs.

The flexibility and relief afforded by the ACGME Stage 3 Pandemic Emergency Status designation allowed us to meet the needs of the surge of patients that required care while we maintained our educational framework and tenets of providing adequate resources and training, working within the confines of safe work hours, ensuring proper supervision, and granting attending privileges to fellows in their core specialty. ●

The 2020 pandemic of coronavirus disease 2019 (COVID-19) has presented significant challenges to the health care workforce.^1,2 As New York City and its environs became the epicenter of the pandemic in the United States, we continued to care for our patients while simultaneously maintaining the education and well-being of our residents.³ Keeping this balance significantly strained resources and presented new challenges for education and service in residency education. What first emerged as an acute emergency has become a chronic disruption in the clinical learning environment. Programs are working to respond to the critical patient needs while ensuring continued progress toward training goals.

Since pregnancy is one condition for which healthy patients continued to require both outpatient visits and inpatient hospitalization, volume was not anticipated to be significantly decreased on our units. Thus, our ObGyn residency programs sought to expeditiously restructure our workforce and educational methods to address the demands of the pandemic. We were aided in our efforts by the Accreditation Council for Graduate Medical Education (ACGME) Extraordinary Circumstances policy. Our institutions were deemed to be functioning at Stage 3 Pandemic Emergency Status, a state in which “the increase in volume and/or severity of illness creates an extraordinary circumstance where routine care, education, and delivery must be reconfigured to focus only on patient care.”⁴

As of May 18, 2020, 26% of residency and fellowship programs in the United States were under Stage 3 COVID-19 Pandemic Emergency Status.⁵ Accordingly, our patient care delivery and educational processes were reconfigured within the context of Stage 3 Status, governed by the overriding principles of ensuring appropriate resources and training, adhering to work hour limits, providing adequate supervision, and credentialing fellows to function in our core specialty.

As ObGyn education leaders from 5 academic medical centers within the COVID-19 epicenter, we present a summary of best practices, based on our experiences, for each of the 4 categories of Stage 3 Status outlined by the ACGME. In an era of globalization, we must learn from pandemics, a call made after the Ebola outbreak in 2015.⁶ We recognize that this type of disruption could happen again with a possible second wave of COVID-19 or another emerging disease.⁷ Thus, we emphasize “lessons learned” that are applicable to a wide range of residency training programs facing various clinical crises.

Ensuring adequate resources and training

Within the context of Stage 3 Status, residency programs have the flexibility to increase residents’ availability in the clinical care setting. However, programs must ensure the safety of both patients and residents.

Continue to: Measures to decrease risk of infection...

One critical resource needed to protect patients and residents is personal protective equipment (PPE). Online instruction and in-person training were used to educate residents and staff on appropriate techniques for donning, doffing, and conserving PPE. Surgical teams were limited to 1 surgeon and 1 resident in each case. In an effort to limit direct contact with COVID-19 infected patients, the number of health care providers rounding on inpatients was restricted, and phone or video conversations were used for communication.

The workforce was modified to decrease exposure to infection and maintain a reserve of healthy residents who were working from home—anticipating that some residents would become ill and this reserve would be called for duty. Similar to other specialties, our programs organized the workforce by arranging residents into teams in which residents worked a number of shifts in a row.^8-12 Regular block schedules were disrupted and non-core rotations were deferred.

As surgeries were canceled and outpatient visits curtailed, many rotations required less resident coverage. Residents were reassigned from rotations where clinical work was suspended to accommodate increased staffing needs in other areas, while accounting for residents who were ill or on leave for postexposure quarantine. Typically, residents worked 12-hour shifts for 3 to 6 days followed by several days off or days working remotely. This team-based strategy decreased the number of residents exposed to COVID-19 at one time, provided time for recuperation, encouraged camaraderie, and enabled residents working remotely to coordinate care and participate in telehealth without direct patient contact.

To minimize high-risk exposure of pregnant residents or residents with underlying health conditions, these residents also worked remotely. Similar to other specialties, it was important to determine essential resident duties and enlist assistance from other clinicians, such as fellows, nurse practitioners, physician assistants, and midwives.

To protect residents and patients, maximizing testing of patients for COVID-19 was an important strategy. Based on early experience at 1 center with patients who were initially asymptomatic but later developed symptoms and tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), universal testing was implemented and endorsed by the New York State COVID-19 Maternity Task Force.¹³ Notably, 87.9% of patients who were positive for SARS-CoV-2 at the time of admission had no symptoms of COVID-19 at presentation. Because the asymptomatic carrier rate appears to be high in obstetric patients, testing of patients is paramount.^3,14 Finally, suspending visitation (except for 1 support person) also was instrumental in decreasing the risk of infection to residents.¹³

Resources for residents with COVID-19

This pandemic placed residency program directors in an unusual situation as frontline caregivers for their own residents. It was imperative to track residents with physical symptoms, conduct testing when possible, and follow the course of residents with confirmed or suspected COVID-19. As serious illness and death have been reported among otherwise healthy young people, we ensured that our homebound residents were frequently monitored.¹⁵ At several of our centers, residents with COVID-19 from any program who chose to separate from their families were provided with alternative housing accommodations. In addition, some of our graduate medical education offices identified specific physicians to care for residents with COVID-19 who did not require hospitalization.

Continue to: Deployment to other specialties...

Several hospitals in the United States redeployed residents because of staffing shortages in high-impact settings.¹² It was important for ObGyns to emphasize that the labor and delivery unit functions as the emergency ward for pregnant women, and that ObGyn residents possess skills specific to the care of these patients.

For our departments, we highlighted that external redeployment could adversely affect our workforce restructuring and, ultimately, patient care. We focused efforts on internal deployment or reassignment as much as possible. Some faculty and fellows in nonobstetric subspecialty areas were redirected to provide care on our inpatient obstetric services.

Educating residents

To maintain educational efforts with social distancing, we used videoconferencing to preserve the protected didactic education time that existed for our residents before the pandemic. This regularly scheduled, nonclinical time also was utilized to instruct residents on the rapidly changing clinical guidelines and to disseminate information about new institutional policies and procedures, ensuring that residents were adequately prepared for their new clinical work.

Work hour requirements

The ACGME requires that work hour limitations remain unchanged during Stage 3 Pandemic Emergency Status. As the pandemic presented new challenges and stressors for residents inside and outside the workplace, ensuring adequate time off to rest and recover was critical for maintaining the resident workforce’s health and wellness.

Thus, our workforce restructuring plans accounted for work hour limitations. As detailed above, the restructuring was accomplished by cohorting residents into small teams that remained unchanged for several weeks. Most shifts were limited to 12 hours, residents continued to be assigned at least 1 day off each week, and daily schedules were structured to ensure at least 10 hours off between shifts. Time spent working remotely was included in work hour calculations.

In addition, residents on “jeopardy” who were available for those who needed to be removed from direct patient care were given at least 1 day off per week in which they could not be pulled for clinical duty. Finally, prolonged inpatient assignments were limited; after these assignments, residents were given increased time for rest and recuperation.

Ensuring adequate supervision

The expectation during Stage 3 Pandemic Emergency Status is that residents, with adequate supervision, provide care that is appropriate for their level of training. To adequately and safely supervise residents, faculty needed training to remain well informed about the clinical care of COVID-19 patients. This was accomplished through frequent communication and consultation with colleagues in infectious disease, occupational health, and guidance from national organizations, such as the American College of Obstetricians and Gynecologists and the Centers for Disease Control and Prevention, and information from our state health departments.

Faculty members were trained in safe donning and doffing of PPE and infection control strategies to ensure they could safely oversee and train residents in these practices. Faculty schedules were significantly altered to ensure an adequate workforce and adequate resident supervision. Faculty efforts were focused on areas of critical need—in our case inpatient obstetrics—with a smaller workforce assigned to outpatient services and inpatient gynecology and gynecologic oncology. Many ObGyn subspecialist faculty were redeployed to general ObGyn inpatient units, thus permitting appropriate resident supervision at all times. In the outpatient setting, faculty adjusted to the changing demands and learned to conduct and supervise telehealth visits.

Finally, for those whose residents were deployed to other services (for example, internal medicine, emergency medicine, or critical care), supervision became paramount. We checked in with our deployed residents daily to be sure that their supervision on those services was adequate. Considering the extreme complexity, rapidly changing understanding of the disease, and often tragic patient outcomes, it was essential to ensure appropriate support and supervision on “off service” deployment.

Continue to: Fellows functioning in core specialty...

Fellows functioning in core specialty

Anticipating the increased need for clinicians on the obstetric services, fellows in subspecialty areas were granted emergency privileges to act as attending faculty in the core specialty, supervising residents and providing patient care. On the other hand, some of those fellows, primarily in gynecologic oncology, were externally redeployed out of core specialty to internal medicine and critical care units. Careful consideration of the fellows’ needs for supervision and support in these roles was essential, and similar support measures that were put in place for our residents were offered to fellows.

In conclusion

The COVID-19 pandemic has presented diverse and complex challenges to the entire health care workforce. Because this crisis is widespread and likely will be lengthy, a sustained and organized response is required.¹⁶ We have highlighted unique challenges specific to residency programs and presented collective best practices from our experiences in ObGyn navigating these obstacles, which are applicable to many other programs.

The flexibility and relief afforded by the ACGME Stage 3 Pandemic Emergency Status designation allowed us to meet the needs of the surge of patients that required care while we maintained our educational framework and tenets of providing adequate resources and training, working within the confines of safe work hours, ensuring proper supervision, and granting attending privileges to fellows in their core specialty. ●

The 2020 pandemic of coronavirus disease 2019 (COVID-19) has presented significant challenges to the health care workforce.^1,2 As New York City and its environs became the epicenter of the pandemic in the United States, we continued to care for our patients while simultaneously maintaining the education and well-being of our residents.³ Keeping this balance significantly strained resources and presented new challenges for education and service in residency education. What first emerged as an acute emergency has become a chronic disruption in the clinical learning environment. Programs are working to respond to the critical patient needs while ensuring continued progress toward training goals.

Since pregnancy is one condition for which healthy patients continued to require both outpatient visits and inpatient hospitalization, volume was not anticipated to be significantly decreased on our units. Thus, our ObGyn residency programs sought to expeditiously restructure our workforce and educational methods to address the demands of the pandemic. We were aided in our efforts by the Accreditation Council for Graduate Medical Education (ACGME) Extraordinary Circumstances policy. Our institutions were deemed to be functioning at Stage 3 Pandemic Emergency Status, a state in which “the increase in volume and/or severity of illness creates an extraordinary circumstance where routine care, education, and delivery must be reconfigured to focus only on patient care.”⁴

As of May 18, 2020, 26% of residency and fellowship programs in the United States were under Stage 3 COVID-19 Pandemic Emergency Status.⁵ Accordingly, our patient care delivery and educational processes were reconfigured within the context of Stage 3 Status, governed by the overriding principles of ensuring appropriate resources and training, adhering to work hour limits, providing adequate supervision, and credentialing fellows to function in our core specialty.

As ObGyn education leaders from 5 academic medical centers within the COVID-19 epicenter, we present a summary of best practices, based on our experiences, for each of the 4 categories of Stage 3 Status outlined by the ACGME. In an era of globalization, we must learn from pandemics, a call made after the Ebola outbreak in 2015.⁶ We recognize that this type of disruption could happen again with a possible second wave of COVID-19 or another emerging disease.⁷ Thus, we emphasize “lessons learned” that are applicable to a wide range of residency training programs facing various clinical crises.

Ensuring adequate resources and training

Within the context of Stage 3 Status, residency programs have the flexibility to increase residents’ availability in the clinical care setting. However, programs must ensure the safety of both patients and residents.

Continue to: Measures to decrease risk of infection...

One critical resource needed to protect patients and residents is personal protective equipment (PPE). Online instruction and in-person training were used to educate residents and staff on appropriate techniques for donning, doffing, and conserving PPE. Surgical teams were limited to 1 surgeon and 1 resident in each case. In an effort to limit direct contact with COVID-19 infected patients, the number of health care providers rounding on inpatients was restricted, and phone or video conversations were used for communication.

The workforce was modified to decrease exposure to infection and maintain a reserve of healthy residents who were working from home—anticipating that some residents would become ill and this reserve would be called for duty. Similar to other specialties, our programs organized the workforce by arranging residents into teams in which residents worked a number of shifts in a row.^8-12 Regular block schedules were disrupted and non-core rotations were deferred.

As surgeries were canceled and outpatient visits curtailed, many rotations required less resident coverage. Residents were reassigned from rotations where clinical work was suspended to accommodate increased staffing needs in other areas, while accounting for residents who were ill or on leave for postexposure quarantine. Typically, residents worked 12-hour shifts for 3 to 6 days followed by several days off or days working remotely. This team-based strategy decreased the number of residents exposed to COVID-19 at one time, provided time for recuperation, encouraged camaraderie, and enabled residents working remotely to coordinate care and participate in telehealth without direct patient contact.

To minimize high-risk exposure of pregnant residents or residents with underlying health conditions, these residents also worked remotely. Similar to other specialties, it was important to determine essential resident duties and enlist assistance from other clinicians, such as fellows, nurse practitioners, physician assistants, and midwives.

To protect residents and patients, maximizing testing of patients for COVID-19 was an important strategy. Based on early experience at 1 center with patients who were initially asymptomatic but later developed symptoms and tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), universal testing was implemented and endorsed by the New York State COVID-19 Maternity Task Force.¹³ Notably, 87.9% of patients who were positive for SARS-CoV-2 at the time of admission had no symptoms of COVID-19 at presentation. Because the asymptomatic carrier rate appears to be high in obstetric patients, testing of patients is paramount.^3,14 Finally, suspending visitation (except for 1 support person) also was instrumental in decreasing the risk of infection to residents.¹³

Resources for residents with COVID-19

This pandemic placed residency program directors in an unusual situation as frontline caregivers for their own residents. It was imperative to track residents with physical symptoms, conduct testing when possible, and follow the course of residents with confirmed or suspected COVID-19. As serious illness and death have been reported among otherwise healthy young people, we ensured that our homebound residents were frequently monitored.¹⁵ At several of our centers, residents with COVID-19 from any program who chose to separate from their families were provided with alternative housing accommodations. In addition, some of our graduate medical education offices identified specific physicians to care for residents with COVID-19 who did not require hospitalization.

Continue to: Deployment to other specialties...

Several hospitals in the United States redeployed residents because of staffing shortages in high-impact settings.¹² It was important for ObGyns to emphasize that the labor and delivery unit functions as the emergency ward for pregnant women, and that ObGyn residents possess skills specific to the care of these patients.

For our departments, we highlighted that external redeployment could adversely affect our workforce restructuring and, ultimately, patient care. We focused efforts on internal deployment or reassignment as much as possible. Some faculty and fellows in nonobstetric subspecialty areas were redirected to provide care on our inpatient obstetric services.

Educating residents

To maintain educational efforts with social distancing, we used videoconferencing to preserve the protected didactic education time that existed for our residents before the pandemic. This regularly scheduled, nonclinical time also was utilized to instruct residents on the rapidly changing clinical guidelines and to disseminate information about new institutional policies and procedures, ensuring that residents were adequately prepared for their new clinical work.

Work hour requirements

The ACGME requires that work hour limitations remain unchanged during Stage 3 Pandemic Emergency Status. As the pandemic presented new challenges and stressors for residents inside and outside the workplace, ensuring adequate time off to rest and recover was critical for maintaining the resident workforce’s health and wellness.

Thus, our workforce restructuring plans accounted for work hour limitations. As detailed above, the restructuring was accomplished by cohorting residents into small teams that remained unchanged for several weeks. Most shifts were limited to 12 hours, residents continued to be assigned at least 1 day off each week, and daily schedules were structured to ensure at least 10 hours off between shifts. Time spent working remotely was included in work hour calculations.

In addition, residents on “jeopardy” who were available for those who needed to be removed from direct patient care were given at least 1 day off per week in which they could not be pulled for clinical duty. Finally, prolonged inpatient assignments were limited; after these assignments, residents were given increased time for rest and recuperation.

Ensuring adequate supervision

The expectation during Stage 3 Pandemic Emergency Status is that residents, with adequate supervision, provide care that is appropriate for their level of training. To adequately and safely supervise residents, faculty needed training to remain well informed about the clinical care of COVID-19 patients. This was accomplished through frequent communication and consultation with colleagues in infectious disease, occupational health, and guidance from national organizations, such as the American College of Obstetricians and Gynecologists and the Centers for Disease Control and Prevention, and information from our state health departments.

Faculty members were trained in safe donning and doffing of PPE and infection control strategies to ensure they could safely oversee and train residents in these practices. Faculty schedules were significantly altered to ensure an adequate workforce and adequate resident supervision. Faculty efforts were focused on areas of critical need—in our case inpatient obstetrics—with a smaller workforce assigned to outpatient services and inpatient gynecology and gynecologic oncology. Many ObGyn subspecialist faculty were redeployed to general ObGyn inpatient units, thus permitting appropriate resident supervision at all times. In the outpatient setting, faculty adjusted to the changing demands and learned to conduct and supervise telehealth visits.

Finally, for those whose residents were deployed to other services (for example, internal medicine, emergency medicine, or critical care), supervision became paramount. We checked in with our deployed residents daily to be sure that their supervision on those services was adequate. Considering the extreme complexity, rapidly changing understanding of the disease, and often tragic patient outcomes, it was essential to ensure appropriate support and supervision on “off service” deployment.

Continue to: Fellows functioning in core specialty...

Fellows functioning in core specialty

Anticipating the increased need for clinicians on the obstetric services, fellows in subspecialty areas were granted emergency privileges to act as attending faculty in the core specialty, supervising residents and providing patient care. On the other hand, some of those fellows, primarily in gynecologic oncology, were externally redeployed out of core specialty to internal medicine and critical care units. Careful consideration of the fellows’ needs for supervision and support in these roles was essential, and similar support measures that were put in place for our residents were offered to fellows.

In conclusion

The COVID-19 pandemic has presented diverse and complex challenges to the entire health care workforce. Because this crisis is widespread and likely will be lengthy, a sustained and organized response is required.¹⁶ We have highlighted unique challenges specific to residency programs and presented collective best practices from our experiences in ObGyn navigating these obstacles, which are applicable to many other programs.

The flexibility and relief afforded by the ACGME Stage 3 Pandemic Emergency Status designation allowed us to meet the needs of the surge of patients that required care while we maintained our educational framework and tenets of providing adequate resources and training, working within the confines of safe work hours, ensuring proper supervision, and granting attending privileges to fellows in their core specialty. ●

CASE Woman with bleeding in early pregnancy

A 31-year-old woman (G1P0) presents to the local emergency department (ED) due to bleeding in pregnancy. She reports a prior open appendectomy for ruptured appendix; she denies a history of sexually transmitted infections, smoking, and contraception use. She reports having regular menstrual cycles and trying to conceive with her husband for 18 months without success until now.

The patient reports that the previous week she took a home pregnancy test that was positive; she endorses having dark brown spotting for the past 2 days but denies pain. Based on the date of her last menstrual period, gestational age is estimated to be 5 weeks and 1 day. Her human chorionic gonadotropin (hCG) level is 1,670 mIU/mL. Transvaginal ultrasonography demonstrates a normal uterus with an endometrial thickness of 10 mm, no evidence of an intrauterine pregnancy (IUP), normal adnexa bilaterally, and scant free fluid in the pelvis.

Identifying and evaluating pregnancy of unknown location

A pregnancy of unknown location (PUL) is defined by a positive serum hCG level in the absence of a visualized IUP or ectopic pregnancy (EP) by pelvic ultrasonography.

Because of variations in screening tools and clinical practices between institutions and care settings (for example, EDs versus specialized outpatient offices), the incidence of PUL is difficult to capture. In specialized early pregnancy clinics, the rate is 8% to 10%, whereas in the ED setting, the PUL rate has been reported to be as high as 42%.^1-6 While approximately 98% to 99% of all pregnancies are intrauterine, only 30% of PULs will continue to develop as viable ongoing intrauterine gestations.^7-9 The remainder are revealed as failing IUPs or EPs. To counsel patients, set expectations, and triage to appropriate management, it is critical to diagnose pregnancy location as efficiently as possible.

Ectopic pregnancy

Ectopic pregnancies represent only 1% to 2% of conceptions (both spontaneous and through assisted reproduction) and occur most commonly in the fallopian tube, although EPs also can implant in the cornua of the uterus, the cervix, cesarean scar, and more rarely on the ovary or abdominal viscera.^10,11 Least common, heterotopic pregnancies—in which an IUP coexists with an EP—occur in 1 in 4,000 to 30,000 pregnancies, more commonly in women who used assisted reproduction.¹¹

Major risk factors for EP include a history of tubal surgery, sexually transmitted infections (particularly Chlamydia trachomatis), pelvic inflammatory disease, conception with an intrauterine device in situ, and a history of prior EP or tubal surgery, particularly prior tubal ligation; minor risk factors include a history of infertility (excluding known tubal factor infertility) or smoking (in a dose-dependent manner).^11,12 The concern for an EP is heightened in patients with these risk factors.

Because of the possibility of rupture and life-threatening hemorrhage, EP carries a risk of significant morbidity and mortality.¹³ Ruptured EPs account for approximately 2.7% of all maternal deaths each year.¹⁴ When diagnosed sufficiently early in a stable patient, most EPs can be managed medically with methotrexate, a folic acid antagonist.¹⁵ Ectopic pregnancies also may be managed surgically, and emergency surgery is indicated in women with evidence of EP rupture and intraperitoneal bleeding.

Continue to: Intrauterine pregnancy...

Intrauterine pregnancy

While excluding EP is critical, it is equally important to diagnose an IUP as expeditiously as possible to avoid inadvertent, destructive intervention. Diagnosis and management of a PUL can involve endometrial aspiration, which would interrupt an IUP and should be avoided until the possibility of a viable IUP has been eliminated in desired pregnancies. The inadvertent administration of methotrexate, a known teratogen, to a patient with an undiagnosed viable IUP can result in miscarriage, elective termination, or a live-born infant with significant malformations, all of which expose the administering physician to malpractice litigation.^16,17

In desired pregnancies, it is essential to differentiate between a viable IUP, a nonviable IUP, and an EP to guide appropriate management and ensure patient safety, whereas exclusion of EP is the priority in undesired pregnancies.

Tools for diagnosing pregnancy location

For diagnosing pregnancy location, serial hCG measurement, transvaginal pelvic ultrasonography, and outpatient endometrial aspiration are all relevant clinical tools. Pregnancy location can be diagnosed with either direct visualization of an IUP or EP by ultrasonography or with confirmed pathology (chorionic villi or trophoblast cells) from endometrial aspiration (FIGURE). A decline in hCG to an undetectable level following endometrial aspiration also is considered sufficient to diagnose a failed IUP, even in the absence of a confirmatory ultrasonography.

Trending hCG values

In stable patients with PUL, serum hCG levels are commonly measured at 2-day intervals, ideally for a minimum of 3 values. Conventional wisdom dictates that in viable IUPs, the hCG level should roughly double every 2 days. However, more recent data suggest that the threshold for minimum expected hCG rise for an ongoing IUP should be far lower when the pregnancy is desired.¹⁸ A less conservative cutoff can be considered when a pregnancy is not desired.

In a multisite cohort study of 1,005 women with PUL, a minimum hCG rise of 35% in 2 days captured the majority of IUPs, with a negative predictive value of 97.2% for IUP.¹⁹ Of note, although the cutoff of 35% was selected to reduce the risk of misdiagnosing an IUP as an EP, 7.7% of IUPs (and 16.8% of EPs) were still misclassified, showing that hCG trends must be interpreted in the context of other clinical data, including ultrasonography findings and patient symptoms and history.

A follow-up study demonstrated that hCG rises are lower (but still within this normal range) when the initial hCG value is higher, particularly greater than 3,000 mIU/mL.²⁰

Studies show that the rate of spontaneous hCG decline in failing IUPs ranges from 12% to 47% in 2 days, falling more rapidly from higher starting hCG values.^19,21 In a retrospective review of 443 women with spontaneously resolving PUL (presumed to be failing IUPs), the minimum 2-day decline in hCG was 35%.²² Any spontaneous hCG decline less than 35% in 2 days in a PUL should raise physician concern for EP.

Conversely, EPs do not demonstrate predictable hCG trends and can mimic the hCG trends of viable or failing IUPs. Although typically half of EPs present with an increasing hCG value and half present with a decreasing hCG value, the majority (71%) demonstrate a slower rate of change than either a viable IUP or a miscarriage.¹¹ This slower change (plateau) should heighten the clinician’s suspicion for an EP.

Continue to: Progesterone levels...

Progesterone levels

A progesterone level often is used to attempt to determine pregnancy viability in women who are not receiving progesterone supplementation, although it ultimately has limited utility. While far less sensitive than an hCG value trend, a serum progesterone level of less than 5 to 10 ng/mL is a rough marker of nonviable pregnancy.²³

In a large meta-analysis of women with pain and bleeding, 96.8% of pregnancies with a single progesterone level of less than 10 ng/mL were nonviable.²³ When an inconclusive ultrasonography was documented in addition to symptoms of pain and bleeding, 99.2% of pregnancies with a progesterone level of less than 3.2 to 6 ng/mL were nonviable.

Progesterone’s usefulness in assessing for a PUL is limited: While progesterone levels may indicate nonviability, they provide no indication of pregnancy location (intrauterine or ectopic).

Alternative serologic markers

Various other reproductive and pregnancy-related hormones have been investigated for use in the diagnosis of pregnancy location in PULs, including activin A, inhibin A, pregnancy-associated plasma protein A (PAPP-A), placental-like growth factor, vascular endothelial growth factor, follistatin, and various microRNAs.^24,25 While research into these biomarkers is ongoing, none have been studied in prospective trials, and they are not for use in current clinical care.

Pelvic ultrasonography

Pelvic ultrasonography is a crucial part of PUL assessment. Transvaginal ultrasonography should be interpreted in the context of the estimated gestational age of the pregnancy and serial hCG values, if available; the patient’s symptoms; and the sensitivity of the ultrasonography equipment, which also may be affected by variables that can reduce visualization, such as uterine fibroids and obesity.

The “discriminatory zone” refers to the hCG value above which an IUP should be visualized by ultrasonography. Generally, with an hCG value of 1,500 to 2,000 mIU/mL or greater, an IUP is expected to be seen with transvaginal sonography.^3,26 Many exceptions to the discriminatory zone have been reported, however, including multiple pregnancies, which will have a higher hCG value at an earlier gestational age. Even in singleton pregnancies, viable IUPs have been documented as developing from PULs with an elevated initial hCG value as high as 4,300 mIU/mL.²⁷ The discriminatory zone may vary among clinical hCG assays, and it also is affected by the quality and modernity of the ultrasonography equipment as well as by the ultrasonography operator’s experience and skill.^28,29

The estimated gestational age, based on either the last menstrual period or assisted reproduction procedure, provides a helpful data point to guide expectations for ultrasonography findings.³⁰ Using transvaginal ultrasonography in a normally progressing IUP, a gestational sac—typically measuring 2 to 3 mm—should be visualized at 5 weeks.^15,30 At approximately 5.5 weeks, a yolk sac measuring 3 to 5 mm should appear. At 6 weeks, an embryo with cardiac activity should be visualized.

In a pregnancy reliably dated beyond 5 weeks, the lack of an intrauterine gestational sac is suspicious for, but not diagnostic of, an EP. Conversely, the visualization of a gestational sac alone (without a yolk sac) is insufficient to definitively exclude an EP, since a small fluid collection in the endometrium (a “pseudosac”) can convincingly mimic the appearance of a gestational sac, and a follow-up ultrasonography should be performed in such cases.

Among patients without ultrasonographic evidence of an IUP, endometrial thickness has been posited as a way to differentiate between IUP and EP.^31,32 Evidence suggests that an endometrial stripe of at least 8 to 10 mm may be somewhat predictive of an IUP, while endometrial thickness below 8 mm is more concerning for EP. This clinical variable, however, has been shown repeatedly to lack sufficient sensitivity and specificity for IUP and should be considered only within the entire clinical context.

Continue to: Endometrial aspiration...

Endometrial aspiration

A persistently abnormal hCG trend and an ultrasonography without evidence of an IUP—particularly with an hCG value above the discriminatory zone and/or with reliable pregnancy dating beyond 5 to 6 weeks—is highly concerning for either a failing IUP or an EP. Once a viable desired IUP is excluded beyond reasonable doubt through these measures, endometrial aspiration to determine pregnancy location is a reasonable next step in PUL management.

Endometrial aspiration can identify a failing IUP by detection of trophoblasts or chorionic villi on pathology and/or by a decline of at least 15% in hCG, measured on the day of endometrial aspiration and again the following day. Endometrial aspiration is effective even in clinical care settings that do not have rapid pathologic analysis available, as hCG measurement before and within 24 hours after the procedure still can be performed.

Vacuum aspiration (electric or manual) in an operating room or office setting is an effective tool for diagnosing pregnancy location.^33,34 The use of an endometrial Pipelle for endometrial sampling (typically used for an office endometrial biopsy to diagnose hyperplasia or malignancy) is insufficient for determining pregnancy location.³⁵ For all patients managed with this protocol, the hCG value ideally should be followed until it is undetectable, regardless of whether an EP or failing IUP was diagnosed. In rare cases, an EP may be diagnosed by a late plateau in hCG values, following an initial decline consistent with a failing IUP.

Utility for diagnosis. Retrospective studies in patients with PUL following in vitro fertilization have established the utility of outpatient endometrial aspiration with a Karman cannula, followed by a repeat hCG measurement on the day after the procedure.^34,36 These data demonstrate that between 42% and 69% of women were ultimately diagnosed with a failed IUP following endometrial aspiration, thereby sparing them unnecessary exposure to methotrexate.

A decline in hCG levels of at least 15% within 24 hours after the procedure indicates that a failed IUP is the most likely diagnosis and further intervention is not indicated (although falling hCG values should be monitored for confirmation); confirmatory pathology with chorionic villi or trophoblasts was present in less than half of these women and is not necessary to diagnose a failed IUP.³⁶ Women diagnosed with a failed IUP after endometrial aspiration also benefitted from a shorter time to resolution of the nonviable pregnancy by approximately 2 weeks.³⁶

Despite the efficacy of endometrial aspiration for the diagnosis of pregnancy location, recent data show that physicians have highly variable approaches to PUL with an hCG plateaued above the discriminatory zone: One-third would first perform endometrial aspiration, while one-third would give methotrexate without further diagnostics.³⁷ Academic physicians were 4 times more likely to recommend endometrial aspiration.³⁷

Presumed EP. Following endometrial aspiration, if pathology does not confirm an intrauterine gestation and the hCG fails to decline by at least 15%, the diagnosis of a presumed EP is made.

For stable patients with neither evidence of intra-abdominal bleeding nor contraindications to methotrexate (such as blood dyscrasias, hepatic or renal insufficiency, active pulmonary or peptic ulcer disease, breastfeeding, or a known intolerance to the medication), methotrexate is recommended for medical management.²⁶ Following screening blood work that includes a complete blood count and liver function and renal function tests, the typical methotrexate dose is 50 mg/m² of body surface area. The single-dose regimen entails checking hCG on the day of methotrexate administration and again on days 4 and 7 thereafter. A minimum decline in hCG of 15% between days 4 and 7 indicates successful treatment; if the hCG decline is below 15%, the patient should receive an additional dose of methotrexate.

There are several published alternative regimens for methotrexate administration, including 2-dose and multidose regimens; the 2-dose protocol (2 doses within 7 days) may be more effective in women with higher hCG (> 3,000 mIU/mL) or known adnexal mass.^26,38

Continue to: Contraindications to methotrexate...

Contraindications to methotrexate. In addition to strict medical contraindications to methotrexate, relative contraindications that indicate a higher risk of methotrexate failure include the presence of fetal cardiac activity, EP mass greater than 4 cm, and serum hCG above 5,000 mIU/mL.²⁶ Because of the potential risk of tubal rupture during medical management, relative contraindications also include patient inability to follow up as an outpatient and patient refusal of blood transfusion.²⁶ Patients with contraindications to methotrexate, hemodynamic instability, ultrasonographic or clinical evidence of EP rupture, or those electing for surgical management may be managed with laparoscopy.¹¹ Discussion of surgical management of EP is beyond the scope of this article.

Follow the hCG level. In patients with a failing IUP or an EP treated with methotrexate or salpingostomy, the hCG level should always be followed until it is negative, usually by weekly measurements once the diagnosis is made. In some cases, the hCG level may plateau after an initial decline, alerting the clinician to failed treatment for a known EP or the need for recategorization of a failed IUP as an EP.

CASE Concluded

The patient’s second and third hCG measurements at 2-day intervals were 1,903 mIU/mL (14% rise) and 2,264 mIU/mL (16% rise). At that point, a repeat transvaginal ultrasonography showed no IUP, adnexal mass, or free fluid. The patient was counseled for outpatient endometrial aspiration, which was performed using manual vacuum aspiration. The serum hCG level on the morning of the procedure was 2,420 mIU/mL. On postprocedure day 1, the serum hCG level fell to 1,615 mIU/mL, a 33% decline. The patient was counseled that this decline in hCG indicated a failing IUP. The final pathologic analysis was returned 3 days later, showing no evidence of trophoblasts and chorionic villi. Regardless, the diagnosis of failing IUP remained given the rapid hCG decline; the tissue from the disrupted failing IUP was likely very scant or simply not drawn into the cannula. Serum hCG levels repeated at weekly intervals revealed ongoing decline, and after 4 weeks, the serum hCG was negative.

In summary

For women diagnosed with PUL, the primary goal is to distinguish an IUP from an EP to reduce the risk of EP rupture through expeditious diagnosis and treatment. In women for whom the pregnancy is desired, distinguishing a viable IUP from a nonviable IUP or an EP is the more specific goal to avoid intervention on a viable IUP (with methotrexate or endometrial aspiration). In women with abnormal hCG trends and indeterminate ultrasonography results (particularly with a serum hCG above the discriminatory zone), outpatient endometrial aspiration is a highly effective way to determine pregnancy location, which dictates further treatment. ●

CASE Woman with bleeding in early pregnancy

A 31-year-old woman (G1P0) presents to the local emergency department (ED) due to bleeding in pregnancy. She reports a prior open appendectomy for ruptured appendix; she denies a history of sexually transmitted infections, smoking, and contraception use. She reports having regular menstrual cycles and trying to conceive with her husband for 18 months without success until now.

The patient reports that the previous week she took a home pregnancy test that was positive; she endorses having dark brown spotting for the past 2 days but denies pain. Based on the date of her last menstrual period, gestational age is estimated to be 5 weeks and 1 day. Her human chorionic gonadotropin (hCG) level is 1,670 mIU/mL. Transvaginal ultrasonography demonstrates a normal uterus with an endometrial thickness of 10 mm, no evidence of an intrauterine pregnancy (IUP), normal adnexa bilaterally, and scant free fluid in the pelvis.

Identifying and evaluating pregnancy of unknown location

A pregnancy of unknown location (PUL) is defined by a positive serum hCG level in the absence of a visualized IUP or ectopic pregnancy (EP) by pelvic ultrasonography.

Because of variations in screening tools and clinical practices between institutions and care settings (for example, EDs versus specialized outpatient offices), the incidence of PUL is difficult to capture. In specialized early pregnancy clinics, the rate is 8% to 10%, whereas in the ED setting, the PUL rate has been reported to be as high as 42%.^1-6 While approximately 98% to 99% of all pregnancies are intrauterine, only 30% of PULs will continue to develop as viable ongoing intrauterine gestations.^7-9 The remainder are revealed as failing IUPs or EPs. To counsel patients, set expectations, and triage to appropriate management, it is critical to diagnose pregnancy location as efficiently as possible.

Ectopic pregnancy

Ectopic pregnancies represent only 1% to 2% of conceptions (both spontaneous and through assisted reproduction) and occur most commonly in the fallopian tube, although EPs also can implant in the cornua of the uterus, the cervix, cesarean scar, and more rarely on the ovary or abdominal viscera.^10,11 Least common, heterotopic pregnancies—in which an IUP coexists with an EP—occur in 1 in 4,000 to 30,000 pregnancies, more commonly in women who used assisted reproduction.¹¹

Major risk factors for EP include a history of tubal surgery, sexually transmitted infections (particularly Chlamydia trachomatis), pelvic inflammatory disease, conception with an intrauterine device in situ, and a history of prior EP or tubal surgery, particularly prior tubal ligation; minor risk factors include a history of infertility (excluding known tubal factor infertility) or smoking (in a dose-dependent manner).^11,12 The concern for an EP is heightened in patients with these risk factors.

Because of the possibility of rupture and life-threatening hemorrhage, EP carries a risk of significant morbidity and mortality.¹³ Ruptured EPs account for approximately 2.7% of all maternal deaths each year.¹⁴ When diagnosed sufficiently early in a stable patient, most EPs can be managed medically with methotrexate, a folic acid antagonist.¹⁵ Ectopic pregnancies also may be managed surgically, and emergency surgery is indicated in women with evidence of EP rupture and intraperitoneal bleeding.

Continue to: Intrauterine pregnancy...

Intrauterine pregnancy

While excluding EP is critical, it is equally important to diagnose an IUP as expeditiously as possible to avoid inadvertent, destructive intervention. Diagnosis and management of a PUL can involve endometrial aspiration, which would interrupt an IUP and should be avoided until the possibility of a viable IUP has been eliminated in desired pregnancies. The inadvertent administration of methotrexate, a known teratogen, to a patient with an undiagnosed viable IUP can result in miscarriage, elective termination, or a live-born infant with significant malformations, all of which expose the administering physician to malpractice litigation.^16,17

In desired pregnancies, it is essential to differentiate between a viable IUP, a nonviable IUP, and an EP to guide appropriate management and ensure patient safety, whereas exclusion of EP is the priority in undesired pregnancies.

Tools for diagnosing pregnancy location

For diagnosing pregnancy location, serial hCG measurement, transvaginal pelvic ultrasonography, and outpatient endometrial aspiration are all relevant clinical tools. Pregnancy location can be diagnosed with either direct visualization of an IUP or EP by ultrasonography or with confirmed pathology (chorionic villi or trophoblast cells) from endometrial aspiration (FIGURE). A decline in hCG to an undetectable level following endometrial aspiration also is considered sufficient to diagnose a failed IUP, even in the absence of a confirmatory ultrasonography.

Trending hCG values

In stable patients with PUL, serum hCG levels are commonly measured at 2-day intervals, ideally for a minimum of 3 values. Conventional wisdom dictates that in viable IUPs, the hCG level should roughly double every 2 days. However, more recent data suggest that the threshold for minimum expected hCG rise for an ongoing IUP should be far lower when the pregnancy is desired.¹⁸ A less conservative cutoff can be considered when a pregnancy is not desired.

In a multisite cohort study of 1,005 women with PUL, a minimum hCG rise of 35% in 2 days captured the majority of IUPs, with a negative predictive value of 97.2% for IUP.¹⁹ Of note, although the cutoff of 35% was selected to reduce the risk of misdiagnosing an IUP as an EP, 7.7% of IUPs (and 16.8% of EPs) were still misclassified, showing that hCG trends must be interpreted in the context of other clinical data, including ultrasonography findings and patient symptoms and history.

A follow-up study demonstrated that hCG rises are lower (but still within this normal range) when the initial hCG value is higher, particularly greater than 3,000 mIU/mL.²⁰

Studies show that the rate of spontaneous hCG decline in failing IUPs ranges from 12% to 47% in 2 days, falling more rapidly from higher starting hCG values.^19,21 In a retrospective review of 443 women with spontaneously resolving PUL (presumed to be failing IUPs), the minimum 2-day decline in hCG was 35%.²² Any spontaneous hCG decline less than 35% in 2 days in a PUL should raise physician concern for EP.

Conversely, EPs do not demonstrate predictable hCG trends and can mimic the hCG trends of viable or failing IUPs. Although typically half of EPs present with an increasing hCG value and half present with a decreasing hCG value, the majority (71%) demonstrate a slower rate of change than either a viable IUP or a miscarriage.¹¹ This slower change (plateau) should heighten the clinician’s suspicion for an EP.

Continue to: Progesterone levels...

Progesterone levels

A progesterone level often is used to attempt to determine pregnancy viability in women who are not receiving progesterone supplementation, although it ultimately has limited utility. While far less sensitive than an hCG value trend, a serum progesterone level of less than 5 to 10 ng/mL is a rough marker of nonviable pregnancy.²³

In a large meta-analysis of women with pain and bleeding, 96.8% of pregnancies with a single progesterone level of less than 10 ng/mL were nonviable.²³ When an inconclusive ultrasonography was documented in addition to symptoms of pain and bleeding, 99.2% of pregnancies with a progesterone level of less than 3.2 to 6 ng/mL were nonviable.

Progesterone’s usefulness in assessing for a PUL is limited: While progesterone levels may indicate nonviability, they provide no indication of pregnancy location (intrauterine or ectopic).

Alternative serologic markers

Various other reproductive and pregnancy-related hormones have been investigated for use in the diagnosis of pregnancy location in PULs, including activin A, inhibin A, pregnancy-associated plasma protein A (PAPP-A), placental-like growth factor, vascular endothelial growth factor, follistatin, and various microRNAs.^24,25 While research into these biomarkers is ongoing, none have been studied in prospective trials, and they are not for use in current clinical care.

Pelvic ultrasonography

Pelvic ultrasonography is a crucial part of PUL assessment. Transvaginal ultrasonography should be interpreted in the context of the estimated gestational age of the pregnancy and serial hCG values, if available; the patient’s symptoms; and the sensitivity of the ultrasonography equipment, which also may be affected by variables that can reduce visualization, such as uterine fibroids and obesity.

The “discriminatory zone” refers to the hCG value above which an IUP should be visualized by ultrasonography. Generally, with an hCG value of 1,500 to 2,000 mIU/mL or greater, an IUP is expected to be seen with transvaginal sonography.^3,26 Many exceptions to the discriminatory zone have been reported, however, including multiple pregnancies, which will have a higher hCG value at an earlier gestational age. Even in singleton pregnancies, viable IUPs have been documented as developing from PULs with an elevated initial hCG value as high as 4,300 mIU/mL.²⁷ The discriminatory zone may vary among clinical hCG assays, and it also is affected by the quality and modernity of the ultrasonography equipment as well as by the ultrasonography operator’s experience and skill.^28,29

The estimated gestational age, based on either the last menstrual period or assisted reproduction procedure, provides a helpful data point to guide expectations for ultrasonography findings.³⁰ Using transvaginal ultrasonography in a normally progressing IUP, a gestational sac—typically measuring 2 to 3 mm—should be visualized at 5 weeks.^15,30 At approximately 5.5 weeks, a yolk sac measuring 3 to 5 mm should appear. At 6 weeks, an embryo with cardiac activity should be visualized.

In a pregnancy reliably dated beyond 5 weeks, the lack of an intrauterine gestational sac is suspicious for, but not diagnostic of, an EP. Conversely, the visualization of a gestational sac alone (without a yolk sac) is insufficient to definitively exclude an EP, since a small fluid collection in the endometrium (a “pseudosac”) can convincingly mimic the appearance of a gestational sac, and a follow-up ultrasonography should be performed in such cases.

Among patients without ultrasonographic evidence of an IUP, endometrial thickness has been posited as a way to differentiate between IUP and EP.^31,32 Evidence suggests that an endometrial stripe of at least 8 to 10 mm may be somewhat predictive of an IUP, while endometrial thickness below 8 mm is more concerning for EP. This clinical variable, however, has been shown repeatedly to lack sufficient sensitivity and specificity for IUP and should be considered only within the entire clinical context.

Continue to: Endometrial aspiration...

Endometrial aspiration

A persistently abnormal hCG trend and an ultrasonography without evidence of an IUP—particularly with an hCG value above the discriminatory zone and/or with reliable pregnancy dating beyond 5 to 6 weeks—is highly concerning for either a failing IUP or an EP. Once a viable desired IUP is excluded beyond reasonable doubt through these measures, endometrial aspiration to determine pregnancy location is a reasonable next step in PUL management.

Endometrial aspiration can identify a failing IUP by detection of trophoblasts or chorionic villi on pathology and/or by a decline of at least 15% in hCG, measured on the day of endometrial aspiration and again the following day. Endometrial aspiration is effective even in clinical care settings that do not have rapid pathologic analysis available, as hCG measurement before and within 24 hours after the procedure still can be performed.

Vacuum aspiration (electric or manual) in an operating room or office setting is an effective tool for diagnosing pregnancy location.^33,34 The use of an endometrial Pipelle for endometrial sampling (typically used for an office endometrial biopsy to diagnose hyperplasia or malignancy) is insufficient for determining pregnancy location.³⁵ For all patients managed with this protocol, the hCG value ideally should be followed until it is undetectable, regardless of whether an EP or failing IUP was diagnosed. In rare cases, an EP may be diagnosed by a late plateau in hCG values, following an initial decline consistent with a failing IUP.

Utility for diagnosis. Retrospective studies in patients with PUL following in vitro fertilization have established the utility of outpatient endometrial aspiration with a Karman cannula, followed by a repeat hCG measurement on the day after the procedure.^34,36 These data demonstrate that between 42% and 69% of women were ultimately diagnosed with a failed IUP following endometrial aspiration, thereby sparing them unnecessary exposure to methotrexate.

A decline in hCG levels of at least 15% within 24 hours after the procedure indicates that a failed IUP is the most likely diagnosis and further intervention is not indicated (although falling hCG values should be monitored for confirmation); confirmatory pathology with chorionic villi or trophoblasts was present in less than half of these women and is not necessary to diagnose a failed IUP.³⁶ Women diagnosed with a failed IUP after endometrial aspiration also benefitted from a shorter time to resolution of the nonviable pregnancy by approximately 2 weeks.³⁶

Despite the efficacy of endometrial aspiration for the diagnosis of pregnancy location, recent data show that physicians have highly variable approaches to PUL with an hCG plateaued above the discriminatory zone: One-third would first perform endometrial aspiration, while one-third would give methotrexate without further diagnostics.³⁷ Academic physicians were 4 times more likely to recommend endometrial aspiration.³⁷

Presumed EP. Following endometrial aspiration, if pathology does not confirm an intrauterine gestation and the hCG fails to decline by at least 15%, the diagnosis of a presumed EP is made.

For stable patients with neither evidence of intra-abdominal bleeding nor contraindications to methotrexate (such as blood dyscrasias, hepatic or renal insufficiency, active pulmonary or peptic ulcer disease, breastfeeding, or a known intolerance to the medication), methotrexate is recommended for medical management.²⁶ Following screening blood work that includes a complete blood count and liver function and renal function tests, the typical methotrexate dose is 50 mg/m² of body surface area. The single-dose regimen entails checking hCG on the day of methotrexate administration and again on days 4 and 7 thereafter. A minimum decline in hCG of 15% between days 4 and 7 indicates successful treatment; if the hCG decline is below 15%, the patient should receive an additional dose of methotrexate.

There are several published alternative regimens for methotrexate administration, including 2-dose and multidose regimens; the 2-dose protocol (2 doses within 7 days) may be more effective in women with higher hCG (> 3,000 mIU/mL) or known adnexal mass.^26,38

Continue to: Contraindications to methotrexate...

Contraindications to methotrexate. In addition to strict medical contraindications to methotrexate, relative contraindications that indicate a higher risk of methotrexate failure include the presence of fetal cardiac activity, EP mass greater than 4 cm, and serum hCG above 5,000 mIU/mL.²⁶ Because of the potential risk of tubal rupture during medical management, relative contraindications also include patient inability to follow up as an outpatient and patient refusal of blood transfusion.²⁶ Patients with contraindications to methotrexate, hemodynamic instability, ultrasonographic or clinical evidence of EP rupture, or those electing for surgical management may be managed with laparoscopy.¹¹ Discussion of surgical management of EP is beyond the scope of this article.

Follow the hCG level. In patients with a failing IUP or an EP treated with methotrexate or salpingostomy, the hCG level should always be followed until it is negative, usually by weekly measurements once the diagnosis is made. In some cases, the hCG level may plateau after an initial decline, alerting the clinician to failed treatment for a known EP or the need for recategorization of a failed IUP as an EP.

CASE Concluded

The patient’s second and third hCG measurements at 2-day intervals were 1,903 mIU/mL (14% rise) and 2,264 mIU/mL (16% rise). At that point, a repeat transvaginal ultrasonography showed no IUP, adnexal mass, or free fluid. The patient was counseled for outpatient endometrial aspiration, which was performed using manual vacuum aspiration. The serum hCG level on the morning of the procedure was 2,420 mIU/mL. On postprocedure day 1, the serum hCG level fell to 1,615 mIU/mL, a 33% decline. The patient was counseled that this decline in hCG indicated a failing IUP. The final pathologic analysis was returned 3 days later, showing no evidence of trophoblasts and chorionic villi. Regardless, the diagnosis of failing IUP remained given the rapid hCG decline; the tissue from the disrupted failing IUP was likely very scant or simply not drawn into the cannula. Serum hCG levels repeated at weekly intervals revealed ongoing decline, and after 4 weeks, the serum hCG was negative.

In summary

For women diagnosed with PUL, the primary goal is to distinguish an IUP from an EP to reduce the risk of EP rupture through expeditious diagnosis and treatment. In women for whom the pregnancy is desired, distinguishing a viable IUP from a nonviable IUP or an EP is the more specific goal to avoid intervention on a viable IUP (with methotrexate or endometrial aspiration). In women with abnormal hCG trends and indeterminate ultrasonography results (particularly with a serum hCG above the discriminatory zone), outpatient endometrial aspiration is a highly effective way to determine pregnancy location, which dictates further treatment. ●

CASE Woman with bleeding in early pregnancy

A 31-year-old woman (G1P0) presents to the local emergency department (ED) due to bleeding in pregnancy. She reports a prior open appendectomy for ruptured appendix; she denies a history of sexually transmitted infections, smoking, and contraception use. She reports having regular menstrual cycles and trying to conceive with her husband for 18 months without success until now.

The patient reports that the previous week she took a home pregnancy test that was positive; she endorses having dark brown spotting for the past 2 days but denies pain. Based on the date of her last menstrual period, gestational age is estimated to be 5 weeks and 1 day. Her human chorionic gonadotropin (hCG) level is 1,670 mIU/mL. Transvaginal ultrasonography demonstrates a normal uterus with an endometrial thickness of 10 mm, no evidence of an intrauterine pregnancy (IUP), normal adnexa bilaterally, and scant free fluid in the pelvis.

Identifying and evaluating pregnancy of unknown location

A pregnancy of unknown location (PUL) is defined by a positive serum hCG level in the absence of a visualized IUP or ectopic pregnancy (EP) by pelvic ultrasonography.

Because of variations in screening tools and clinical practices between institutions and care settings (for example, EDs versus specialized outpatient offices), the incidence of PUL is difficult to capture. In specialized early pregnancy clinics, the rate is 8% to 10%, whereas in the ED setting, the PUL rate has been reported to be as high as 42%.^1-6 While approximately 98% to 99% of all pregnancies are intrauterine, only 30% of PULs will continue to develop as viable ongoing intrauterine gestations.^7-9 The remainder are revealed as failing IUPs or EPs. To counsel patients, set expectations, and triage to appropriate management, it is critical to diagnose pregnancy location as efficiently as possible.

Ectopic pregnancy

Ectopic pregnancies represent only 1% to 2% of conceptions (both spontaneous and through assisted reproduction) and occur most commonly in the fallopian tube, although EPs also can implant in the cornua of the uterus, the cervix, cesarean scar, and more rarely on the ovary or abdominal viscera.^10,11 Least common, heterotopic pregnancies—in which an IUP coexists with an EP—occur in 1 in 4,000 to 30,000 pregnancies, more commonly in women who used assisted reproduction.¹¹

Major risk factors for EP include a history of tubal surgery, sexually transmitted infections (particularly Chlamydia trachomatis), pelvic inflammatory disease, conception with an intrauterine device in situ, and a history of prior EP or tubal surgery, particularly prior tubal ligation; minor risk factors include a history of infertility (excluding known tubal factor infertility) or smoking (in a dose-dependent manner).^11,12 The concern for an EP is heightened in patients with these risk factors.

Because of the possibility of rupture and life-threatening hemorrhage, EP carries a risk of significant morbidity and mortality.¹³ Ruptured EPs account for approximately 2.7% of all maternal deaths each year.¹⁴ When diagnosed sufficiently early in a stable patient, most EPs can be managed medically with methotrexate, a folic acid antagonist.¹⁵ Ectopic pregnancies also may be managed surgically, and emergency surgery is indicated in women with evidence of EP rupture and intraperitoneal bleeding.

Continue to: Intrauterine pregnancy...

Intrauterine pregnancy

While excluding EP is critical, it is equally important to diagnose an IUP as expeditiously as possible to avoid inadvertent, destructive intervention. Diagnosis and management of a PUL can involve endometrial aspiration, which would interrupt an IUP and should be avoided until the possibility of a viable IUP has been eliminated in desired pregnancies. The inadvertent administration of methotrexate, a known teratogen, to a patient with an undiagnosed viable IUP can result in miscarriage, elective termination, or a live-born infant with significant malformations, all of which expose the administering physician to malpractice litigation.^16,17

In desired pregnancies, it is essential to differentiate between a viable IUP, a nonviable IUP, and an EP to guide appropriate management and ensure patient safety, whereas exclusion of EP is the priority in undesired pregnancies.

Tools for diagnosing pregnancy location

For diagnosing pregnancy location, serial hCG measurement, transvaginal pelvic ultrasonography, and outpatient endometrial aspiration are all relevant clinical tools. Pregnancy location can be diagnosed with either direct visualization of an IUP or EP by ultrasonography or with confirmed pathology (chorionic villi or trophoblast cells) from endometrial aspiration (FIGURE). A decline in hCG to an undetectable level following endometrial aspiration also is considered sufficient to diagnose a failed IUP, even in the absence of a confirmatory ultrasonography.

Trending hCG values

In stable patients with PUL, serum hCG levels are commonly measured at 2-day intervals, ideally for a minimum of 3 values. Conventional wisdom dictates that in viable IUPs, the hCG level should roughly double every 2 days. However, more recent data suggest that the threshold for minimum expected hCG rise for an ongoing IUP should be far lower when the pregnancy is desired.¹⁸ A less conservative cutoff can be considered when a pregnancy is not desired.

In a multisite cohort study of 1,005 women with PUL, a minimum hCG rise of 35% in 2 days captured the majority of IUPs, with a negative predictive value of 97.2% for IUP.¹⁹ Of note, although the cutoff of 35% was selected to reduce the risk of misdiagnosing an IUP as an EP, 7.7% of IUPs (and 16.8% of EPs) were still misclassified, showing that hCG trends must be interpreted in the context of other clinical data, including ultrasonography findings and patient symptoms and history.

A follow-up study demonstrated that hCG rises are lower (but still within this normal range) when the initial hCG value is higher, particularly greater than 3,000 mIU/mL.²⁰

Studies show that the rate of spontaneous hCG decline in failing IUPs ranges from 12% to 47% in 2 days, falling more rapidly from higher starting hCG values.^19,21 In a retrospective review of 443 women with spontaneously resolving PUL (presumed to be failing IUPs), the minimum 2-day decline in hCG was 35%.²² Any spontaneous hCG decline less than 35% in 2 days in a PUL should raise physician concern for EP.

Conversely, EPs do not demonstrate predictable hCG trends and can mimic the hCG trends of viable or failing IUPs. Although typically half of EPs present with an increasing hCG value and half present with a decreasing hCG value, the majority (71%) demonstrate a slower rate of change than either a viable IUP or a miscarriage.¹¹ This slower change (plateau) should heighten the clinician’s suspicion for an EP.

Continue to: Progesterone levels...

Progesterone levels

A progesterone level often is used to attempt to determine pregnancy viability in women who are not receiving progesterone supplementation, although it ultimately has limited utility. While far less sensitive than an hCG value trend, a serum progesterone level of less than 5 to 10 ng/mL is a rough marker of nonviable pregnancy.²³

In a large meta-analysis of women with pain and bleeding, 96.8% of pregnancies with a single progesterone level of less than 10 ng/mL were nonviable.²³ When an inconclusive ultrasonography was documented in addition to symptoms of pain and bleeding, 99.2% of pregnancies with a progesterone level of less than 3.2 to 6 ng/mL were nonviable.

Progesterone’s usefulness in assessing for a PUL is limited: While progesterone levels may indicate nonviability, they provide no indication of pregnancy location (intrauterine or ectopic).

Alternative serologic markers

Various other reproductive and pregnancy-related hormones have been investigated for use in the diagnosis of pregnancy location in PULs, including activin A, inhibin A, pregnancy-associated plasma protein A (PAPP-A), placental-like growth factor, vascular endothelial growth factor, follistatin, and various microRNAs.^24,25 While research into these biomarkers is ongoing, none have been studied in prospective trials, and they are not for use in current clinical care.

Pelvic ultrasonography

Pelvic ultrasonography is a crucial part of PUL assessment. Transvaginal ultrasonography should be interpreted in the context of the estimated gestational age of the pregnancy and serial hCG values, if available; the patient’s symptoms; and the sensitivity of the ultrasonography equipment, which also may be affected by variables that can reduce visualization, such as uterine fibroids and obesity.

The “discriminatory zone” refers to the hCG value above which an IUP should be visualized by ultrasonography. Generally, with an hCG value of 1,500 to 2,000 mIU/mL or greater, an IUP is expected to be seen with transvaginal sonography.^3,26 Many exceptions to the discriminatory zone have been reported, however, including multiple pregnancies, which will have a higher hCG value at an earlier gestational age. Even in singleton pregnancies, viable IUPs have been documented as developing from PULs with an elevated initial hCG value as high as 4,300 mIU/mL.²⁷ The discriminatory zone may vary among clinical hCG assays, and it also is affected by the quality and modernity of the ultrasonography equipment as well as by the ultrasonography operator’s experience and skill.^28,29

The estimated gestational age, based on either the last menstrual period or assisted reproduction procedure, provides a helpful data point to guide expectations for ultrasonography findings.³⁰ Using transvaginal ultrasonography in a normally progressing IUP, a gestational sac—typically measuring 2 to 3 mm—should be visualized at 5 weeks.^15,30 At approximately 5.5 weeks, a yolk sac measuring 3 to 5 mm should appear. At 6 weeks, an embryo with cardiac activity should be visualized.

In a pregnancy reliably dated beyond 5 weeks, the lack of an intrauterine gestational sac is suspicious for, but not diagnostic of, an EP. Conversely, the visualization of a gestational sac alone (without a yolk sac) is insufficient to definitively exclude an EP, since a small fluid collection in the endometrium (a “pseudosac”) can convincingly mimic the appearance of a gestational sac, and a follow-up ultrasonography should be performed in such cases.

Among patients without ultrasonographic evidence of an IUP, endometrial thickness has been posited as a way to differentiate between IUP and EP.^31,32 Evidence suggests that an endometrial stripe of at least 8 to 10 mm may be somewhat predictive of an IUP, while endometrial thickness below 8 mm is more concerning for EP. This clinical variable, however, has been shown repeatedly to lack sufficient sensitivity and specificity for IUP and should be considered only within the entire clinical context.

Continue to: Endometrial aspiration...

Endometrial aspiration

A persistently abnormal hCG trend and an ultrasonography without evidence of an IUP—particularly with an hCG value above the discriminatory zone and/or with reliable pregnancy dating beyond 5 to 6 weeks—is highly concerning for either a failing IUP or an EP. Once a viable desired IUP is excluded beyond reasonable doubt through these measures, endometrial aspiration to determine pregnancy location is a reasonable next step in PUL management.

Endometrial aspiration can identify a failing IUP by detection of trophoblasts or chorionic villi on pathology and/or by a decline of at least 15% in hCG, measured on the day of endometrial aspiration and again the following day. Endometrial aspiration is effective even in clinical care settings that do not have rapid pathologic analysis available, as hCG measurement before and within 24 hours after the procedure still can be performed.

Vacuum aspiration (electric or manual) in an operating room or office setting is an effective tool for diagnosing pregnancy location.^33,34 The use of an endometrial Pipelle for endometrial sampling (typically used for an office endometrial biopsy to diagnose hyperplasia or malignancy) is insufficient for determining pregnancy location.³⁵ For all patients managed with this protocol, the hCG value ideally should be followed until it is undetectable, regardless of whether an EP or failing IUP was diagnosed. In rare cases, an EP may be diagnosed by a late plateau in hCG values, following an initial decline consistent with a failing IUP.

Utility for diagnosis. Retrospective studies in patients with PUL following in vitro fertilization have established the utility of outpatient endometrial aspiration with a Karman cannula, followed by a repeat hCG measurement on the day after the procedure.^34,36 These data demonstrate that between 42% and 69% of women were ultimately diagnosed with a failed IUP following endometrial aspiration, thereby sparing them unnecessary exposure to methotrexate.

A decline in hCG levels of at least 15% within 24 hours after the procedure indicates that a failed IUP is the most likely diagnosis and further intervention is not indicated (although falling hCG values should be monitored for confirmation); confirmatory pathology with chorionic villi or trophoblasts was present in less than half of these women and is not necessary to diagnose a failed IUP.³⁶ Women diagnosed with a failed IUP after endometrial aspiration also benefitted from a shorter time to resolution of the nonviable pregnancy by approximately 2 weeks.³⁶

Despite the efficacy of endometrial aspiration for the diagnosis of pregnancy location, recent data show that physicians have highly variable approaches to PUL with an hCG plateaued above the discriminatory zone: One-third would first perform endometrial aspiration, while one-third would give methotrexate without further diagnostics.³⁷ Academic physicians were 4 times more likely to recommend endometrial aspiration.³⁷

Presumed EP. Following endometrial aspiration, if pathology does not confirm an intrauterine gestation and the hCG fails to decline by at least 15%, the diagnosis of a presumed EP is made.

For stable patients with neither evidence of intra-abdominal bleeding nor contraindications to methotrexate (such as blood dyscrasias, hepatic or renal insufficiency, active pulmonary or peptic ulcer disease, breastfeeding, or a known intolerance to the medication), methotrexate is recommended for medical management.²⁶ Following screening blood work that includes a complete blood count and liver function and renal function tests, the typical methotrexate dose is 50 mg/m² of body surface area. The single-dose regimen entails checking hCG on the day of methotrexate administration and again on days 4 and 7 thereafter. A minimum decline in hCG of 15% between days 4 and 7 indicates successful treatment; if the hCG decline is below 15%, the patient should receive an additional dose of methotrexate.

There are several published alternative regimens for methotrexate administration, including 2-dose and multidose regimens; the 2-dose protocol (2 doses within 7 days) may be more effective in women with higher hCG (> 3,000 mIU/mL) or known adnexal mass.^26,38

Continue to: Contraindications to methotrexate...

Contraindications to methotrexate. In addition to strict medical contraindications to methotrexate, relative contraindications that indicate a higher risk of methotrexate failure include the presence of fetal cardiac activity, EP mass greater than 4 cm, and serum hCG above 5,000 mIU/mL.²⁶ Because of the potential risk of tubal rupture during medical management, relative contraindications also include patient inability to follow up as an outpatient and patient refusal of blood transfusion.²⁶ Patients with contraindications to methotrexate, hemodynamic instability, ultrasonographic or clinical evidence of EP rupture, or those electing for surgical management may be managed with laparoscopy.¹¹ Discussion of surgical management of EP is beyond the scope of this article.

Follow the hCG level. In patients with a failing IUP or an EP treated with methotrexate or salpingostomy, the hCG level should always be followed until it is negative, usually by weekly measurements once the diagnosis is made. In some cases, the hCG level may plateau after an initial decline, alerting the clinician to failed treatment for a known EP or the need for recategorization of a failed IUP as an EP.

CASE Concluded

The patient’s second and third hCG measurements at 2-day intervals were 1,903 mIU/mL (14% rise) and 2,264 mIU/mL (16% rise). At that point, a repeat transvaginal ultrasonography showed no IUP, adnexal mass, or free fluid. The patient was counseled for outpatient endometrial aspiration, which was performed using manual vacuum aspiration. The serum hCG level on the morning of the procedure was 2,420 mIU/mL. On postprocedure day 1, the serum hCG level fell to 1,615 mIU/mL, a 33% decline. The patient was counseled that this decline in hCG indicated a failing IUP. The final pathologic analysis was returned 3 days later, showing no evidence of trophoblasts and chorionic villi. Regardless, the diagnosis of failing IUP remained given the rapid hCG decline; the tissue from the disrupted failing IUP was likely very scant or simply not drawn into the cannula. Serum hCG levels repeated at weekly intervals revealed ongoing decline, and after 4 weeks, the serum hCG was negative.

In summary

For women diagnosed with PUL, the primary goal is to distinguish an IUP from an EP to reduce the risk of EP rupture through expeditious diagnosis and treatment. In women for whom the pregnancy is desired, distinguishing a viable IUP from a nonviable IUP or an EP is the more specific goal to avoid intervention on a viable IUP (with methotrexate or endometrial aspiration). In women with abnormal hCG trends and indeterminate ultrasonography results (particularly with a serum hCG above the discriminatory zone), outpatient endometrial aspiration is a highly effective way to determine pregnancy location, which dictates further treatment. ●

The advent of biologic therapy over the last 2 decades has transformed the treatment of psoriasis; patients who either are not good candidates for or have an inadequate response to traditional treatments (topicals and/or phototherapy) now have numerous options for treatment.¹ Patients burdened by extensive disease, recurrent flares, and stubborn treatment areas are ideal candidates for biologics. There are 11 biologics approved by the US Food and Drug Administration (FDA)(Table) for treating moderate to severe plaque psoriasis as supported by grade A evidence. The FDA has authorized 1 new biologic—risankizumab—since the joint guidelines from the American Academy of Dermatology and National Psoriasis Foundation were released for the treatment of psoriasis with biologics.² This article aims to address updates on recent clinical trial findings (April 2019 to April 2020) regarding biologic therapy initiation and maintenance for adult patients. Prescribers should use this update as guidance for determining the appropriate biologic class based on patient characteristics and for approaching biologic-experienced patients with refractory psoriasis. This update also may serve as a reference for the recommended dosing regimens of the 11 approved biologics.

Using Risankizumab

Risankizumab is a new biologic that selectively targets the IL-23 pathway by binding the p19 subunit of IL-23. It was approved by the FDA in April 2019. Two recent studies have demonstrated the efficacy of risankizumab in disease management.^3,4

IMMvent was a double-blind, 2-part, phase 3, randomized controlled trial (RCT) of participants 18 years and older (N=605) with moderate to severe psoriasis (with or without psoriatic arthritis) across 11 countries.³ Inclusion criteria consisted of psoriasis involving at least 10% of the body surface area (BSA), absolute psoriasis area and severity index (PASI) score of 12 or higher, and static physician global assessment (sPGA) score of 3 or higher. Prior biologic treatment did not preclude study entry (excluding risankizumab or adalimumab), and nearly 40% of participants previously had been on a different biologic. Notably, this trial allowed for inclusion of patients with prior malignancy (>5 years prior) and patients who tested positive for exposure to tuberculosis (TB) but were not shown to have active TB (provided appropriate treatment for latent TB was started). Study participants identified as white (81%), Asian (14%), black/African American (4%), or other ethnicity (1%). Part A involved administration of 150 mg risankizumab (n=301) at weeks 0 and 4 or 80 mg adalimumab (n=304) loading dose at week 0 followed by 40 mg at week 1 and 40 mg every other week thereafter until the end of week 15. At week 16 there was a significant difference in proportion of participants achieving 90% or more improvement (PASI-90) with risankizumab (72%) vs adalimumab (47%)(P<.0001) and achieving an sPGA score of 0 or 1 (clear or almost clear) with risankizumab (84%) vs adalimumab (60%)(P<.0001). In part B (weeks 16–44), adalimumab immediate responder (PASI ≥50 to PASI <90) participants were re-randomized to continue adalimumab 40 mg every other week (starting from week 17 and stopping at week 44) or switch to 150 mg risankizumab administered at weeks 16, 20, and 32. Patients taking risankizumab in part A continued the drug, administered at weeks 16 and 28. At week 44, there was a significant difference in percentage of participants achieving PASI-90 with risankizumab (66%) vs adalimumab (21%)(P<.0001).³

IMMhance was another double-blind phase 3 RCT with 2 parts that assessed the clinical efficacy of risankizumab compared to placebo in patients 18 years or older (N=507) across 9 countries with the same inclusion criteria for patients as IMMvent.⁴ Part A involved administration of 150 mg risankizumab (n=407) or placebo (n=100) at weeks 0 and 4 using a 4:1 random allocation ratio. At week 16, regardless of initial treatment, all participants received 150 mg risankizumab. Treatment results at week 16 showed a significant difference in percentage of participants achieving PASI-90 with risankizumab (73.2%) vs placebo (2.0%)(P<.001) and sPGA score of 0 or 1 with risankizumab (83.5%) vs placebo (7.0%)(P<.001). Furthermore, in part B (weeks 16–104), at week 28 participants on risankizumab with an sPGA score of 0 or 1 were randomized with a 1:2 allocation ratio to continue 150 mg risankizumab or switch to placebo to produce a treatment withdrawal effect. Part B results showed a significant difference in the proportion of participants achieving an sPGA score of 0 or 1 with risankizumab (87.4%) vs placebo (61.3%)(P<.001) at week 52 and at week 104 with risankizumab (81.1%) vs placebo (7.1%)(P<.001). Risankizumab was well tolerated, with the most common adverse events (AEs) being nasopharyngitis (23.4%), upper respiratory tract infection (15.4%), and headache (6.8%). Serious AEs included cancer (2.6%; 2.2 events per 100 patient-years), hepatic events (4.6%) including hepatic cirrhosis (0.2%), and serious infections (1.8%; 1.4 events per 100 patient-years).⁴

Overall, the strengths of risankizumab with regard to its clinical efficacy and utility in biologic-experienced patients were confirmed in these studies. The inclusion of patients with prior treated malignancy and positive TB tests also was more in line with what one might encounter with real-world practice and, as such, provided valuable data to help aid treatment decisions. These 2 studies provided valuable evidence for the therapeutic benefit and relatively mild safety profile of risankizumab in treatment of moderate to severe psoriasis for patients with and without prior biologic therapy.

Choosing a Biologic

Refractory psoriasis involves nonresponse (primary failure) or return of disease symptoms after initial improvement (secondary failure) with a biologic. Selecting a biologic for patients who have experienced prior biologic failure is difficult. It is still unknown whether it is more efficacious for patients to try a same-class drug or a biologic targeting a different inflammatory pathway or cytokine. Studies have shown mixed results regarding how to manage patients with biologic failure, with both approaches demonstrating positive outcomes.

One analysis of the Corrona Psoriasis Registry included 144 patients, the majority of whom (89.8%) were biologic experienced, who began secukinumab treatment and returned for a 6-month follow-up (5–9 months).⁵ Patients enrolled in the registry were 18 years or older, had been diagnosed with psoriasis by a dermatologist, and initiated or switched an FDA-approved systemic agent or biologic within the last 12 months. Of biologic-experienced participants, 37.7% had used 3 or more biologics. More than half of included participants were either male (55%) or obese (53.4%). Comorbidities included hypertension (43.2%), hyperlipidemia (33.9%), anxiety (20.3%), diabetes mellitus (15.3%), cardiovascular disease (14.4%), and depression (13.6%). After 6 months of treatment, there was significant improvement in the involvement of BSA (mean difference, −12.1), investigator global assessment score (−1.5), dermatology life quality index (DLQI)(−4.8), pain (−23.2), itch (−30.8), fatigue (−8.8), and work productivity (−9.2)(P<.01). Secukinumab therapy displayed notable reduction in symptom severity in this population with difficult-to-treat psoriasis. Its relative success in this cohort provides support for its use in treating patients who have failed other classes of biologics.⁵

Evidence supporting reduction of pruritus and pain with secukinumab also was notable. The CLEAR phase 3 RCT involved participants treated with 300 mg secukinumab every week for the first 4 weeks and then every 4 weeks thereafter for 48 weeks (n=312), up to 100 weeks (n=277).⁶ Participants had complete relief of pain (score 0), itching, and scaling at week 16 (69.4%, 49.7%, and 61.2%, respectively), week 52 (67.1%, 48.9%, and 53.3%, respectively), and week 104 (70.9%, 47.4%, and 54.8%, respectively). Reported AEs included candida infections (7.2%), malignant or unspecified tumors (1.5%), and neutropenia (<1%).⁶

Researchers investigated intraclass switching to brodalumab with prior failure of IL-17 inhibitors. An open-label study involved participants (n=39) with prior failure with secukinumab or ixekizumab therapy.⁷ Participants were administered 210 mg brodalumab with standard dosing at weeks 0, 1, and 2, and then every 2 weeks thereafter. At week 16, 69% of participants achieved PASI-75, 44% achieved PASI-90, 28% achieved PASI-100, and 62% achieved an sPGA score of 0 or 1. The authors attributed the relative success of brodalumab compared to prior anti–IL-17 agents to inhibition of the IL-17 receptor with brodalumab rather than the IL-17A ligand.⁷ Brodalumab may be a useful alternative biologic for patients with nonresponse to and secondary failure with biologics, including the IL-17A inhibitors.

Recent findings support effective skin clearance and improved symptom management with ixekizumab and ustekinumab. Of note, ixekizumab was reported to provide rapid improvement in skin lesions and quality of life to a greater extent than guselkumab.

The IXORA-R double-blinded RCT compared the clinical benefit of participants 18 years and older taking standard approved dosages of ixekizumab (n=520) or guselkumab (n=507).⁸ Patients were included if they had plaque psoriasis for at least 6 months before baseline, an sPGA score of at least 3, PASI score of 12 or higher, 10% or greater BSA, no prior IL-17 inhibitor failure, no use of IL-23 p19 inhibitors, and no use of any biologic within the specified period prior to baseline. At week 12, ixekizumab showed superior clinical improvement measured by the proportion of participants achieving complete skin clearance (ie, PASI-100)(41%) compared to guselkumab (25%)(P<.001). There were more participants taking ixekizumab who reported DLQI of 0 or 1 (no impact of disease on quality of life)(34%) compared to guselkumab (21%)(P<.001) as early on as week 4. The most common AE was upper respiratory tract infection (7%) in both groups. The risk of treatment-emergent AEs (56%), discontinuation because of AEs (2%), and serious AEs (3%) were comparable in both groups. The number of injection-site reactions was higher with ixekizumab (13%) vs guselkumab (3%). The authors concluded that ixekizumab offers the ability to provide rapid relief of symptoms, which is associated with improved DLQI.⁸

Response to ustekinumab therapy was assessed in a patient cohort enrolled in the Corrona Psoriasis Registry. This study involved 178 participants 18 years and older with psoriasis involvement of 3% or greater BSA who were treated with ustekinumab.⁹ By their 6-month follow-up visit, 55.6% of participants achieved adequate treatment response (BSA improving to <3% or 75% from enrollment). Increasing patient age was significantly associated with decreased likelihood of achieving a response (odds ratio, 0.981 [95% confidence interval, 0.962-0.999]; P=.049). Ustekinumab is a practical option for psoriasis treatment that seems to yield better results in younger patients.⁹ This evidence reveals that increased patient age is a characteristic that may contribute to poor treatment response and should be considered when choosing the best fit for biologic therapy.

Final Thoughts

Using evidence-based interventions to treat patients is the cornerstone of ethical and high-quality medical care. This guide sought to provide relevant updates in a variety of both comparator and pivotal trials, with the goal of summarizing clinically relevant information that may be extracted from these trials to guide patient care. It is not an exhaustive review but may be utilized as a reference tool to fine-tune selection criteria in choosing 1 of 11 biologics for the treatment of psoriasis.

The advent of biologic therapy over the last 2 decades has transformed the treatment of psoriasis; patients who either are not good candidates for or have an inadequate response to traditional treatments (topicals and/or phototherapy) now have numerous options for treatment.¹ Patients burdened by extensive disease, recurrent flares, and stubborn treatment areas are ideal candidates for biologics. There are 11 biologics approved by the US Food and Drug Administration (FDA)(Table) for treating moderate to severe plaque psoriasis as supported by grade A evidence. The FDA has authorized 1 new biologic—risankizumab—since the joint guidelines from the American Academy of Dermatology and National Psoriasis Foundation were released for the treatment of psoriasis with biologics.² This article aims to address updates on recent clinical trial findings (April 2019 to April 2020) regarding biologic therapy initiation and maintenance for adult patients. Prescribers should use this update as guidance for determining the appropriate biologic class based on patient characteristics and for approaching biologic-experienced patients with refractory psoriasis. This update also may serve as a reference for the recommended dosing regimens of the 11 approved biologics.

Using Risankizumab

Risankizumab is a new biologic that selectively targets the IL-23 pathway by binding the p19 subunit of IL-23. It was approved by the FDA in April 2019. Two recent studies have demonstrated the efficacy of risankizumab in disease management.^3,4

IMMvent was a double-blind, 2-part, phase 3, randomized controlled trial (RCT) of participants 18 years and older (N=605) with moderate to severe psoriasis (with or without psoriatic arthritis) across 11 countries.³ Inclusion criteria consisted of psoriasis involving at least 10% of the body surface area (BSA), absolute psoriasis area and severity index (PASI) score of 12 or higher, and static physician global assessment (sPGA) score of 3 or higher. Prior biologic treatment did not preclude study entry (excluding risankizumab or adalimumab), and nearly 40% of participants previously had been on a different biologic. Notably, this trial allowed for inclusion of patients with prior malignancy (>5 years prior) and patients who tested positive for exposure to tuberculosis (TB) but were not shown to have active TB (provided appropriate treatment for latent TB was started). Study participants identified as white (81%), Asian (14%), black/African American (4%), or other ethnicity (1%). Part A involved administration of 150 mg risankizumab (n=301) at weeks 0 and 4 or 80 mg adalimumab (n=304) loading dose at week 0 followed by 40 mg at week 1 and 40 mg every other week thereafter until the end of week 15. At week 16 there was a significant difference in proportion of participants achieving 90% or more improvement (PASI-90) with risankizumab (72%) vs adalimumab (47%)(P<.0001) and achieving an sPGA score of 0 or 1 (clear or almost clear) with risankizumab (84%) vs adalimumab (60%)(P<.0001). In part B (weeks 16–44), adalimumab immediate responder (PASI ≥50 to PASI <90) participants were re-randomized to continue adalimumab 40 mg every other week (starting from week 17 and stopping at week 44) or switch to 150 mg risankizumab administered at weeks 16, 20, and 32. Patients taking risankizumab in part A continued the drug, administered at weeks 16 and 28. At week 44, there was a significant difference in percentage of participants achieving PASI-90 with risankizumab (66%) vs adalimumab (21%)(P<.0001).³

IMMhance was another double-blind phase 3 RCT with 2 parts that assessed the clinical efficacy of risankizumab compared to placebo in patients 18 years or older (N=507) across 9 countries with the same inclusion criteria for patients as IMMvent.⁴ Part A involved administration of 150 mg risankizumab (n=407) or placebo (n=100) at weeks 0 and 4 using a 4:1 random allocation ratio. At week 16, regardless of initial treatment, all participants received 150 mg risankizumab. Treatment results at week 16 showed a significant difference in percentage of participants achieving PASI-90 with risankizumab (73.2%) vs placebo (2.0%)(P<.001) and sPGA score of 0 or 1 with risankizumab (83.5%) vs placebo (7.0%)(P<.001). Furthermore, in part B (weeks 16–104), at week 28 participants on risankizumab with an sPGA score of 0 or 1 were randomized with a 1:2 allocation ratio to continue 150 mg risankizumab or switch to placebo to produce a treatment withdrawal effect. Part B results showed a significant difference in the proportion of participants achieving an sPGA score of 0 or 1 with risankizumab (87.4%) vs placebo (61.3%)(P<.001) at week 52 and at week 104 with risankizumab (81.1%) vs placebo (7.1%)(P<.001). Risankizumab was well tolerated, with the most common adverse events (AEs) being nasopharyngitis (23.4%), upper respiratory tract infection (15.4%), and headache (6.8%). Serious AEs included cancer (2.6%; 2.2 events per 100 patient-years), hepatic events (4.6%) including hepatic cirrhosis (0.2%), and serious infections (1.8%; 1.4 events per 100 patient-years).⁴

Overall, the strengths of risankizumab with regard to its clinical efficacy and utility in biologic-experienced patients were confirmed in these studies. The inclusion of patients with prior treated malignancy and positive TB tests also was more in line with what one might encounter with real-world practice and, as such, provided valuable data to help aid treatment decisions. These 2 studies provided valuable evidence for the therapeutic benefit and relatively mild safety profile of risankizumab in treatment of moderate to severe psoriasis for patients with and without prior biologic therapy.

Choosing a Biologic

Refractory psoriasis involves nonresponse (primary failure) or return of disease symptoms after initial improvement (secondary failure) with a biologic. Selecting a biologic for patients who have experienced prior biologic failure is difficult. It is still unknown whether it is more efficacious for patients to try a same-class drug or a biologic targeting a different inflammatory pathway or cytokine. Studies have shown mixed results regarding how to manage patients with biologic failure, with both approaches demonstrating positive outcomes.

One analysis of the Corrona Psoriasis Registry included 144 patients, the majority of whom (89.8%) were biologic experienced, who began secukinumab treatment and returned for a 6-month follow-up (5–9 months).⁵ Patients enrolled in the registry were 18 years or older, had been diagnosed with psoriasis by a dermatologist, and initiated or switched an FDA-approved systemic agent or biologic within the last 12 months. Of biologic-experienced participants, 37.7% had used 3 or more biologics. More than half of included participants were either male (55%) or obese (53.4%). Comorbidities included hypertension (43.2%), hyperlipidemia (33.9%), anxiety (20.3%), diabetes mellitus (15.3%), cardiovascular disease (14.4%), and depression (13.6%). After 6 months of treatment, there was significant improvement in the involvement of BSA (mean difference, −12.1), investigator global assessment score (−1.5), dermatology life quality index (DLQI)(−4.8), pain (−23.2), itch (−30.8), fatigue (−8.8), and work productivity (−9.2)(P<.01). Secukinumab therapy displayed notable reduction in symptom severity in this population with difficult-to-treat psoriasis. Its relative success in this cohort provides support for its use in treating patients who have failed other classes of biologics.⁵

Evidence supporting reduction of pruritus and pain with secukinumab also was notable. The CLEAR phase 3 RCT involved participants treated with 300 mg secukinumab every week for the first 4 weeks and then every 4 weeks thereafter for 48 weeks (n=312), up to 100 weeks (n=277).⁶ Participants had complete relief of pain (score 0), itching, and scaling at week 16 (69.4%, 49.7%, and 61.2%, respectively), week 52 (67.1%, 48.9%, and 53.3%, respectively), and week 104 (70.9%, 47.4%, and 54.8%, respectively). Reported AEs included candida infections (7.2%), malignant or unspecified tumors (1.5%), and neutropenia (<1%).⁶

Researchers investigated intraclass switching to brodalumab with prior failure of IL-17 inhibitors. An open-label study involved participants (n=39) with prior failure with secukinumab or ixekizumab therapy.⁷ Participants were administered 210 mg brodalumab with standard dosing at weeks 0, 1, and 2, and then every 2 weeks thereafter. At week 16, 69% of participants achieved PASI-75, 44% achieved PASI-90, 28% achieved PASI-100, and 62% achieved an sPGA score of 0 or 1. The authors attributed the relative success of brodalumab compared to prior anti–IL-17 agents to inhibition of the IL-17 receptor with brodalumab rather than the IL-17A ligand.⁷ Brodalumab may be a useful alternative biologic for patients with nonresponse to and secondary failure with biologics, including the IL-17A inhibitors.

Recent findings support effective skin clearance and improved symptom management with ixekizumab and ustekinumab. Of note, ixekizumab was reported to provide rapid improvement in skin lesions and quality of life to a greater extent than guselkumab.

The IXORA-R double-blinded RCT compared the clinical benefit of participants 18 years and older taking standard approved dosages of ixekizumab (n=520) or guselkumab (n=507).⁸ Patients were included if they had plaque psoriasis for at least 6 months before baseline, an sPGA score of at least 3, PASI score of 12 or higher, 10% or greater BSA, no prior IL-17 inhibitor failure, no use of IL-23 p19 inhibitors, and no use of any biologic within the specified period prior to baseline. At week 12, ixekizumab showed superior clinical improvement measured by the proportion of participants achieving complete skin clearance (ie, PASI-100)(41%) compared to guselkumab (25%)(P<.001). There were more participants taking ixekizumab who reported DLQI of 0 or 1 (no impact of disease on quality of life)(34%) compared to guselkumab (21%)(P<.001) as early on as week 4. The most common AE was upper respiratory tract infection (7%) in both groups. The risk of treatment-emergent AEs (56%), discontinuation because of AEs (2%), and serious AEs (3%) were comparable in both groups. The number of injection-site reactions was higher with ixekizumab (13%) vs guselkumab (3%). The authors concluded that ixekizumab offers the ability to provide rapid relief of symptoms, which is associated with improved DLQI.⁸

Response to ustekinumab therapy was assessed in a patient cohort enrolled in the Corrona Psoriasis Registry. This study involved 178 participants 18 years and older with psoriasis involvement of 3% or greater BSA who were treated with ustekinumab.⁹ By their 6-month follow-up visit, 55.6% of participants achieved adequate treatment response (BSA improving to <3% or 75% from enrollment). Increasing patient age was significantly associated with decreased likelihood of achieving a response (odds ratio, 0.981 [95% confidence interval, 0.962-0.999]; P=.049). Ustekinumab is a practical option for psoriasis treatment that seems to yield better results in younger patients.⁹ This evidence reveals that increased patient age is a characteristic that may contribute to poor treatment response and should be considered when choosing the best fit for biologic therapy.

Final Thoughts

Using evidence-based interventions to treat patients is the cornerstone of ethical and high-quality medical care. This guide sought to provide relevant updates in a variety of both comparator and pivotal trials, with the goal of summarizing clinically relevant information that may be extracted from these trials to guide patient care. It is not an exhaustive review but may be utilized as a reference tool to fine-tune selection criteria in choosing 1 of 11 biologics for the treatment of psoriasis.

The advent of biologic therapy over the last 2 decades has transformed the treatment of psoriasis; patients who either are not good candidates for or have an inadequate response to traditional treatments (topicals and/or phototherapy) now have numerous options for treatment.¹ Patients burdened by extensive disease, recurrent flares, and stubborn treatment areas are ideal candidates for biologics. There are 11 biologics approved by the US Food and Drug Administration (FDA)(Table) for treating moderate to severe plaque psoriasis as supported by grade A evidence. The FDA has authorized 1 new biologic—risankizumab—since the joint guidelines from the American Academy of Dermatology and National Psoriasis Foundation were released for the treatment of psoriasis with biologics.² This article aims to address updates on recent clinical trial findings (April 2019 to April 2020) regarding biologic therapy initiation and maintenance for adult patients. Prescribers should use this update as guidance for determining the appropriate biologic class based on patient characteristics and for approaching biologic-experienced patients with refractory psoriasis. This update also may serve as a reference for the recommended dosing regimens of the 11 approved biologics.

Using Risankizumab

Risankizumab is a new biologic that selectively targets the IL-23 pathway by binding the p19 subunit of IL-23. It was approved by the FDA in April 2019. Two recent studies have demonstrated the efficacy of risankizumab in disease management.^3,4

IMMvent was a double-blind, 2-part, phase 3, randomized controlled trial (RCT) of participants 18 years and older (N=605) with moderate to severe psoriasis (with or without psoriatic arthritis) across 11 countries.³ Inclusion criteria consisted of psoriasis involving at least 10% of the body surface area (BSA), absolute psoriasis area and severity index (PASI) score of 12 or higher, and static physician global assessment (sPGA) score of 3 or higher. Prior biologic treatment did not preclude study entry (excluding risankizumab or adalimumab), and nearly 40% of participants previously had been on a different biologic. Notably, this trial allowed for inclusion of patients with prior malignancy (>5 years prior) and patients who tested positive for exposure to tuberculosis (TB) but were not shown to have active TB (provided appropriate treatment for latent TB was started). Study participants identified as white (81%), Asian (14%), black/African American (4%), or other ethnicity (1%). Part A involved administration of 150 mg risankizumab (n=301) at weeks 0 and 4 or 80 mg adalimumab (n=304) loading dose at week 0 followed by 40 mg at week 1 and 40 mg every other week thereafter until the end of week 15. At week 16 there was a significant difference in proportion of participants achieving 90% or more improvement (PASI-90) with risankizumab (72%) vs adalimumab (47%)(P<.0001) and achieving an sPGA score of 0 or 1 (clear or almost clear) with risankizumab (84%) vs adalimumab (60%)(P<.0001). In part B (weeks 16–44), adalimumab immediate responder (PASI ≥50 to PASI <90) participants were re-randomized to continue adalimumab 40 mg every other week (starting from week 17 and stopping at week 44) or switch to 150 mg risankizumab administered at weeks 16, 20, and 32. Patients taking risankizumab in part A continued the drug, administered at weeks 16 and 28. At week 44, there was a significant difference in percentage of participants achieving PASI-90 with risankizumab (66%) vs adalimumab (21%)(P<.0001).³

IMMhance was another double-blind phase 3 RCT with 2 parts that assessed the clinical efficacy of risankizumab compared to placebo in patients 18 years or older (N=507) across 9 countries with the same inclusion criteria for patients as IMMvent.⁴ Part A involved administration of 150 mg risankizumab (n=407) or placebo (n=100) at weeks 0 and 4 using a 4:1 random allocation ratio. At week 16, regardless of initial treatment, all participants received 150 mg risankizumab. Treatment results at week 16 showed a significant difference in percentage of participants achieving PASI-90 with risankizumab (73.2%) vs placebo (2.0%)(P<.001) and sPGA score of 0 or 1 with risankizumab (83.5%) vs placebo (7.0%)(P<.001). Furthermore, in part B (weeks 16–104), at week 28 participants on risankizumab with an sPGA score of 0 or 1 were randomized with a 1:2 allocation ratio to continue 150 mg risankizumab or switch to placebo to produce a treatment withdrawal effect. Part B results showed a significant difference in the proportion of participants achieving an sPGA score of 0 or 1 with risankizumab (87.4%) vs placebo (61.3%)(P<.001) at week 52 and at week 104 with risankizumab (81.1%) vs placebo (7.1%)(P<.001). Risankizumab was well tolerated, with the most common adverse events (AEs) being nasopharyngitis (23.4%), upper respiratory tract infection (15.4%), and headache (6.8%). Serious AEs included cancer (2.6%; 2.2 events per 100 patient-years), hepatic events (4.6%) including hepatic cirrhosis (0.2%), and serious infections (1.8%; 1.4 events per 100 patient-years).⁴

Overall, the strengths of risankizumab with regard to its clinical efficacy and utility in biologic-experienced patients were confirmed in these studies. The inclusion of patients with prior treated malignancy and positive TB tests also was more in line with what one might encounter with real-world practice and, as such, provided valuable data to help aid treatment decisions. These 2 studies provided valuable evidence for the therapeutic benefit and relatively mild safety profile of risankizumab in treatment of moderate to severe psoriasis for patients with and without prior biologic therapy.

Choosing a Biologic

Refractory psoriasis involves nonresponse (primary failure) or return of disease symptoms after initial improvement (secondary failure) with a biologic. Selecting a biologic for patients who have experienced prior biologic failure is difficult. It is still unknown whether it is more efficacious for patients to try a same-class drug or a biologic targeting a different inflammatory pathway or cytokine. Studies have shown mixed results regarding how to manage patients with biologic failure, with both approaches demonstrating positive outcomes.

One analysis of the Corrona Psoriasis Registry included 144 patients, the majority of whom (89.8%) were biologic experienced, who began secukinumab treatment and returned for a 6-month follow-up (5–9 months).⁵ Patients enrolled in the registry were 18 years or older, had been diagnosed with psoriasis by a dermatologist, and initiated or switched an FDA-approved systemic agent or biologic within the last 12 months. Of biologic-experienced participants, 37.7% had used 3 or more biologics. More than half of included participants were either male (55%) or obese (53.4%). Comorbidities included hypertension (43.2%), hyperlipidemia (33.9%), anxiety (20.3%), diabetes mellitus (15.3%), cardiovascular disease (14.4%), and depression (13.6%). After 6 months of treatment, there was significant improvement in the involvement of BSA (mean difference, −12.1), investigator global assessment score (−1.5), dermatology life quality index (DLQI)(−4.8), pain (−23.2), itch (−30.8), fatigue (−8.8), and work productivity (−9.2)(P<.01). Secukinumab therapy displayed notable reduction in symptom severity in this population with difficult-to-treat psoriasis. Its relative success in this cohort provides support for its use in treating patients who have failed other classes of biologics.⁵

Evidence supporting reduction of pruritus and pain with secukinumab also was notable. The CLEAR phase 3 RCT involved participants treated with 300 mg secukinumab every week for the first 4 weeks and then every 4 weeks thereafter for 48 weeks (n=312), up to 100 weeks (n=277).⁶ Participants had complete relief of pain (score 0), itching, and scaling at week 16 (69.4%, 49.7%, and 61.2%, respectively), week 52 (67.1%, 48.9%, and 53.3%, respectively), and week 104 (70.9%, 47.4%, and 54.8%, respectively). Reported AEs included candida infections (7.2%), malignant or unspecified tumors (1.5%), and neutropenia (<1%).⁶

Researchers investigated intraclass switching to brodalumab with prior failure of IL-17 inhibitors. An open-label study involved participants (n=39) with prior failure with secukinumab or ixekizumab therapy.⁷ Participants were administered 210 mg brodalumab with standard dosing at weeks 0, 1, and 2, and then every 2 weeks thereafter. At week 16, 69% of participants achieved PASI-75, 44% achieved PASI-90, 28% achieved PASI-100, and 62% achieved an sPGA score of 0 or 1. The authors attributed the relative success of brodalumab compared to prior anti–IL-17 agents to inhibition of the IL-17 receptor with brodalumab rather than the IL-17A ligand.⁷ Brodalumab may be a useful alternative biologic for patients with nonresponse to and secondary failure with biologics, including the IL-17A inhibitors.

Recent findings support effective skin clearance and improved symptom management with ixekizumab and ustekinumab. Of note, ixekizumab was reported to provide rapid improvement in skin lesions and quality of life to a greater extent than guselkumab.

The IXORA-R double-blinded RCT compared the clinical benefit of participants 18 years and older taking standard approved dosages of ixekizumab (n=520) or guselkumab (n=507).⁸ Patients were included if they had plaque psoriasis for at least 6 months before baseline, an sPGA score of at least 3, PASI score of 12 or higher, 10% or greater BSA, no prior IL-17 inhibitor failure, no use of IL-23 p19 inhibitors, and no use of any biologic within the specified period prior to baseline. At week 12, ixekizumab showed superior clinical improvement measured by the proportion of participants achieving complete skin clearance (ie, PASI-100)(41%) compared to guselkumab (25%)(P<.001). There were more participants taking ixekizumab who reported DLQI of 0 or 1 (no impact of disease on quality of life)(34%) compared to guselkumab (21%)(P<.001) as early on as week 4. The most common AE was upper respiratory tract infection (7%) in both groups. The risk of treatment-emergent AEs (56%), discontinuation because of AEs (2%), and serious AEs (3%) were comparable in both groups. The number of injection-site reactions was higher with ixekizumab (13%) vs guselkumab (3%). The authors concluded that ixekizumab offers the ability to provide rapid relief of symptoms, which is associated with improved DLQI.⁸

Response to ustekinumab therapy was assessed in a patient cohort enrolled in the Corrona Psoriasis Registry. This study involved 178 participants 18 years and older with psoriasis involvement of 3% or greater BSA who were treated with ustekinumab.⁹ By their 6-month follow-up visit, 55.6% of participants achieved adequate treatment response (BSA improving to <3% or 75% from enrollment). Increasing patient age was significantly associated with decreased likelihood of achieving a response (odds ratio, 0.981 [95% confidence interval, 0.962-0.999]; P=.049). Ustekinumab is a practical option for psoriasis treatment that seems to yield better results in younger patients.⁹ This evidence reveals that increased patient age is a characteristic that may contribute to poor treatment response and should be considered when choosing the best fit for biologic therapy.

Final Thoughts

Using evidence-based interventions to treat patients is the cornerstone of ethical and high-quality medical care. This guide sought to provide relevant updates in a variety of both comparator and pivotal trials, with the goal of summarizing clinically relevant information that may be extracted from these trials to guide patient care. It is not an exhaustive review but may be utilized as a reference tool to fine-tune selection criteria in choosing 1 of 11 biologics for the treatment of psoriasis.

Historically, there have been limited data available on the management of psoriasis in pregnancy. The most comprehensive discussion of treatment guidelines is from 2012.¹ In the interim, many biologics have been approved for treating psoriasis, with slow accumulation of pregnancy safety data. The 2019 American Academy of Dermatology–National Psoriasis Foundation guidelines on biologics for psoriasis contain updated information but also highlight the paucity of pregnancy safety data.² This gap is in part a consequence of the exclusion and disenrollment of pregnant women from clinical trials.³ Additionally, lack of detection through registries contributes; pregnancy capture in registries is low compared to the expected number of pregnancies estimated from US Census data.⁴ Despite these shortcomings, psoriasis patients who are already pregnant or are considering becoming pregnant frequently are encountered in practice and may need treatment. This article reviews the evidence on commonly used treatments for psoriasis in pregnancy.

Background

For many patients, psoriasis improves during pregnancy^5,6 and becomes worse postpartum. In a prospective study, most patients reported improvement in pregnancy corresponding to a significant decrease in affected body surface area (P<.001) by 10 to 20 weeks’ gestation. Most patients also reported worsening of psoriasis postpartum; a significant increase in psoriatic body surface area (P=.001) was observed after delivery.⁷ Despite these findings, a considerable number of patients also experience stable disease or worsening of disease during pregnancy.

In addition to the maternal disease state, the issue of pregnancy outcomes is paramount. In the inflammatory bowel disease and rheumatology literature, it is established that uncontrolled disease is associated with poorer pregnancy outcomes.^8-10 Guidelines vary among societies on the use of biologics in pregnancy generally (eTable 1^1,2,9,11-24), but some societies recommend systemic agents to achieve disease control during pregnancy.^9,25

Assessing the potential interplay between disease severity and outcomes in pregnant women with psoriasis is further complicated by the slowly growing body of literature demonstrating that women with psoriasis have more comorbidities²⁶ and worse pregnancy outcomes.^27,28 Pregnant psoriasis patients are more likely to smoke, have depression, and be overweight or obese prior to pregnancy and are less likely to take prenatal vitamins.²⁶ They also have an increased risk for cesarean birth, gestational diabetes, gestational hypertension, and preeclampsia.²⁸ In contrast to these prior studies, a systematic review revealed no risk for adverse outcomes in pregnant women with psoriasis.²⁹

Assessment of Treatments for Psoriasis in Pregnancy

In light of these issues, treatment of psoriasis during pregnancy should be assessed from several vantage points. Of note, the US Food and Drug Administration changed its classification scheme in 2015 to a more narrative format called the Pregnancy and Lactation Labeling Rule.³⁰ Prior classifications, however, provide a reasonable starting point for categorizing the safety of drugs (Table³¹). Importantly, time of exposure to systemic agents also matters; first-trimester exposure is more likely to affect embryogenesis, whereas second- and third-trimester exposures are more prone to affect other aspects of fetal growth. eTable 2 provides data on the use of oral and topical medications to treat psoriasis in pregnancy.^1,8,22,32-45

Topical Agents
Topical steroids are largely understood to be reasonable treatment options, though consideration of potency, formulation, area of application, and use of occlusion is important.^1,46 Risk for orofacial cleft has been noted with first-trimester topical steroid exposure, though a 2015 Cochrane review update determined that the relative risk of this association was not significantly elevated.³²

The impact of topical calcipotriene and salicylic acid has not been studied in human pregnancies,¹ but systemic absorption can occur for both. There is potential for vitamin D toxicity with calcipotriene⁴⁶; consequently, use during pregnancy is not recommended.^1,46 Some authors recommend against topical salicylic acid in pregnancy; others report that limited exposure is permissible.⁴⁷ In fact, as salicylic acid commonly is found in over-the-counter acne products, many women of childbearing potential likely have quotidian exposure.

Preterm delivery and low birthweight have been reported with oral tacrolimus; however, risk with topical tacrolimus is thought to be low¹ because the molecular size likely prohibits notable absorption.⁴⁷ Evidence for the use of anthralin and coal tar also is scarce. First-trimester coal tar use should be avoided; subsequent use in pregnancy should be restricted given concern for adverse outcomes.¹

Phototherapy
Broadband or narrowband UVB therapy is recommended as second-line therapy in pregnancy. No cases of fetal risk or premature delivery associated with UVB therapy were found in our search.¹ Phototherapy can exacerbate melasma⁴⁷ and decrease folate levels⁴⁸; as such, some authors recommend folate supplementation in females of childbearing age who are being treated with phototherapy.⁴⁹ Psoralen, used in psoralen plus UVA therapy, is mutagenic and therefore contraindicated in pregnancy.¹

Oral Medications
Both methotrexate, which is a teratogen, abortifacient, and mutagen,¹ and systemic retinoids, which are teratogens, are contraindicated in pregnancy.^1,47 Acitretin labeling recommends avoiding pregnancy for 3 years posttreatment⁵⁰ because alcohol intake prolongs the medication’s half-life.²²

Apremilast use is not documented in pregnant psoriasis patients⁵¹; an ongoing registry of the Organization of Tetralogy Information Specialists has not reported publicly to date.⁵² Animal studies of apremilast have documented dose-related decreased birthweight and fetal loss.²²

Safety data for systemic steroids, used infrequently in psoriasis, are not well established. First-trimester prednisone exposure has been associated with prematurity, low birthweight, and congenital abnormalities.³⁸ A separate evaluation of 1047 children exposed to betamethasone in utero failed to demonstrate significant change in birthweight or head circumference. However, repeat antenatal corticosteroid exposure was associated with attention problems at 2 years of age.³⁹

Data regarding cyclosporine use, derived primarily from organ transplant recipients, suggest elevated risk for prematurity and low birthweight.^53,54 A meta-analysis demonstrated that organ transplant recipients taking cyclosporine had a nonsignificantly elevated odds ratio for congenital malformations, prematurity, and low birthweight.⁴² Cyclosporine use for psoriasis in pregnancy is not well described; in a study, rates of prematurity and low birthweight were both 21%.⁴³ Limited data are available for Janus kinase inhibitors, none of which are approved for psoriasis, though clinical trials in psoriasis and psoriatic arthritis are underway (ClinicalTrials.gov identifiers NCT04246372, NCT03104374, NCT03104400).

Biologics and Small-Molecule Inhibitors
Limited data on biologics in pregnancy exist²⁵ (eTable 3). Placental transport of IgG antibodies, including biologics, increases throughout pregnancy, especially in the third trimester.⁸² Infants of mothers treated with a biologic with potential for placental transfer are therefore considered by some authors to be immunosuppressed during the first months of life.²

Looking globally across biologics used for psoriasis, limited safety data are encouraging. In a review of PSOLAR (Psoriasis Longitudinal Assessment and Registry), 83 pregnancies with biologic exposure resulted in 59 live births (71%); 18 spontaneous abortions (22%); 6 induced abortions (7%); no congenital abnormalities; and 7 reports of neonatal problems, including respiratory issues, ABO blood group mismatch, hospitalization, and opioid withdrawal.⁸³

Use of tumor necrosis factor (TNF) inhibitors in pregnancy has the most data²⁵ and is considered a reasonable treatment option. Historically, there was concern about the risk for VACTERL syndrome (vertebral defects, anal atresia, cardiac defects, tracheoesophageal fistula, renal anomalies, limb abnormalities) with exposure to a TNF inhibitor,^25,84-86 but further reports have alleviated these concerns. Active transplacental transport occurs for adalimumab, infliximab, and golimumab,⁸⁷ but given structural differences, transport of certolizumab and etanercept is substantially less.^88,89 In the CRIB study of placental transfer of certolizumab from mother to infant (N=14), pharmacokinetic data demonstrated no quantifiable certolizumab levels in 13 infants and minimal levels in 1 infant at birth.⁸⁸ There are fewer data available on the use of other biologics in pregnancy, but for those in which active placental transport is relevant, similar concerns (ie, immunosuppression) might arise (eTable 3).

Concern over biologics largely involves risk for newborn immunosuppression. A case report detailed a Crohn disease patient treated with infliximab who gave birth to an infant who died of disseminated bacille Calmette-Guérin infection at 4.5 months after receiving the vaccine at 3 months.⁹⁰ This case underscores the importance of delaying live vaccination in infants born to mothers who were treated with a biologic during pregnancy. Authors have provided various data on how long to avoid vaccination; some state as long as 1 year.⁹¹

In pregnant females with inflammatory bowel disease treated with a biologic, no correlation was observed among maternal, placental, and infant serum biologic levels and neonatal infection. However, an association between preterm birth and the level of the biologic in maternal and placental (but not infant) serum and preterm birth was observed.⁹²

In another report from the same registry, combination therapy with a TNF inhibitor and another immunomodulator led to an increased risk for infection in infants at 12 months of age, compared to infants exposed to monotherapy⁸⁹ or exposed to neither agent.⁹³ A strategy to circumvent this potential problem is to avoid treatment with actively transported molecules in the third trimester.

Conclusion

Limited data exist to guide providers who are treating pregnant women with psoriasis. Our understanding of treatment of psoriasis in pregnancy is limited as a consequence of regulations surrounding clinical trials and inadequate detection of pregnancies in registries. Further efforts are necessary to better understand the relationship between psoriasis and pregnancy and how to manage pregnant women with psoriasis.

Historically, there have been limited data available on the management of psoriasis in pregnancy. The most comprehensive discussion of treatment guidelines is from 2012.¹ In the interim, many biologics have been approved for treating psoriasis, with slow accumulation of pregnancy safety data. The 2019 American Academy of Dermatology–National Psoriasis Foundation guidelines on biologics for psoriasis contain updated information but also highlight the paucity of pregnancy safety data.² This gap is in part a consequence of the exclusion and disenrollment of pregnant women from clinical trials.³ Additionally, lack of detection through registries contributes; pregnancy capture in registries is low compared to the expected number of pregnancies estimated from US Census data.⁴ Despite these shortcomings, psoriasis patients who are already pregnant or are considering becoming pregnant frequently are encountered in practice and may need treatment. This article reviews the evidence on commonly used treatments for psoriasis in pregnancy.

Background

For many patients, psoriasis improves during pregnancy^5,6 and becomes worse postpartum. In a prospective study, most patients reported improvement in pregnancy corresponding to a significant decrease in affected body surface area (P<.001) by 10 to 20 weeks’ gestation. Most patients also reported worsening of psoriasis postpartum; a significant increase in psoriatic body surface area (P=.001) was observed after delivery.⁷ Despite these findings, a considerable number of patients also experience stable disease or worsening of disease during pregnancy.

In addition to the maternal disease state, the issue of pregnancy outcomes is paramount. In the inflammatory bowel disease and rheumatology literature, it is established that uncontrolled disease is associated with poorer pregnancy outcomes.^8-10 Guidelines vary among societies on the use of biologics in pregnancy generally (eTable 1^1,2,9,11-24), but some societies recommend systemic agents to achieve disease control during pregnancy.^9,25

Assessing the potential interplay between disease severity and outcomes in pregnant women with psoriasis is further complicated by the slowly growing body of literature demonstrating that women with psoriasis have more comorbidities²⁶ and worse pregnancy outcomes.^27,28 Pregnant psoriasis patients are more likely to smoke, have depression, and be overweight or obese prior to pregnancy and are less likely to take prenatal vitamins.²⁶ They also have an increased risk for cesarean birth, gestational diabetes, gestational hypertension, and preeclampsia.²⁸ In contrast to these prior studies, a systematic review revealed no risk for adverse outcomes in pregnant women with psoriasis.²⁹

Assessment of Treatments for Psoriasis in Pregnancy

In light of these issues, treatment of psoriasis during pregnancy should be assessed from several vantage points. Of note, the US Food and Drug Administration changed its classification scheme in 2015 to a more narrative format called the Pregnancy and Lactation Labeling Rule.³⁰ Prior classifications, however, provide a reasonable starting point for categorizing the safety of drugs (Table³¹). Importantly, time of exposure to systemic agents also matters; first-trimester exposure is more likely to affect embryogenesis, whereas second- and third-trimester exposures are more prone to affect other aspects of fetal growth. eTable 2 provides data on the use of oral and topical medications to treat psoriasis in pregnancy.^1,8,22,32-45

Topical Agents
Topical steroids are largely understood to be reasonable treatment options, though consideration of potency, formulation, area of application, and use of occlusion is important.^1,46 Risk for orofacial cleft has been noted with first-trimester topical steroid exposure, though a 2015 Cochrane review update determined that the relative risk of this association was not significantly elevated.³²

The impact of topical calcipotriene and salicylic acid has not been studied in human pregnancies,¹ but systemic absorption can occur for both. There is potential for vitamin D toxicity with calcipotriene⁴⁶; consequently, use during pregnancy is not recommended.^1,46 Some authors recommend against topical salicylic acid in pregnancy; others report that limited exposure is permissible.⁴⁷ In fact, as salicylic acid commonly is found in over-the-counter acne products, many women of childbearing potential likely have quotidian exposure.

Preterm delivery and low birthweight have been reported with oral tacrolimus; however, risk with topical tacrolimus is thought to be low¹ because the molecular size likely prohibits notable absorption.⁴⁷ Evidence for the use of anthralin and coal tar also is scarce. First-trimester coal tar use should be avoided; subsequent use in pregnancy should be restricted given concern for adverse outcomes.¹

Phototherapy
Broadband or narrowband UVB therapy is recommended as second-line therapy in pregnancy. No cases of fetal risk or premature delivery associated with UVB therapy were found in our search.¹ Phototherapy can exacerbate melasma⁴⁷ and decrease folate levels⁴⁸; as such, some authors recommend folate supplementation in females of childbearing age who are being treated with phototherapy.⁴⁹ Psoralen, used in psoralen plus UVA therapy, is mutagenic and therefore contraindicated in pregnancy.¹

Oral Medications
Both methotrexate, which is a teratogen, abortifacient, and mutagen,¹ and systemic retinoids, which are teratogens, are contraindicated in pregnancy.^1,47 Acitretin labeling recommends avoiding pregnancy for 3 years posttreatment⁵⁰ because alcohol intake prolongs the medication’s half-life.²²

Apremilast use is not documented in pregnant psoriasis patients⁵¹; an ongoing registry of the Organization of Tetralogy Information Specialists has not reported publicly to date.⁵² Animal studies of apremilast have documented dose-related decreased birthweight and fetal loss.²²

Safety data for systemic steroids, used infrequently in psoriasis, are not well established. First-trimester prednisone exposure has been associated with prematurity, low birthweight, and congenital abnormalities.³⁸ A separate evaluation of 1047 children exposed to betamethasone in utero failed to demonstrate significant change in birthweight or head circumference. However, repeat antenatal corticosteroid exposure was associated with attention problems at 2 years of age.³⁹

Data regarding cyclosporine use, derived primarily from organ transplant recipients, suggest elevated risk for prematurity and low birthweight.^53,54 A meta-analysis demonstrated that organ transplant recipients taking cyclosporine had a nonsignificantly elevated odds ratio for congenital malformations, prematurity, and low birthweight.⁴² Cyclosporine use for psoriasis in pregnancy is not well described; in a study, rates of prematurity and low birthweight were both 21%.⁴³ Limited data are available for Janus kinase inhibitors, none of which are approved for psoriasis, though clinical trials in psoriasis and psoriatic arthritis are underway (ClinicalTrials.gov identifiers NCT04246372, NCT03104374, NCT03104400).

Biologics and Small-Molecule Inhibitors
Limited data on biologics in pregnancy exist²⁵ (eTable 3). Placental transport of IgG antibodies, including biologics, increases throughout pregnancy, especially in the third trimester.⁸² Infants of mothers treated with a biologic with potential for placental transfer are therefore considered by some authors to be immunosuppressed during the first months of life.²

Looking globally across biologics used for psoriasis, limited safety data are encouraging. In a review of PSOLAR (Psoriasis Longitudinal Assessment and Registry), 83 pregnancies with biologic exposure resulted in 59 live births (71%); 18 spontaneous abortions (22%); 6 induced abortions (7%); no congenital abnormalities; and 7 reports of neonatal problems, including respiratory issues, ABO blood group mismatch, hospitalization, and opioid withdrawal.⁸³

Use of tumor necrosis factor (TNF) inhibitors in pregnancy has the most data²⁵ and is considered a reasonable treatment option. Historically, there was concern about the risk for VACTERL syndrome (vertebral defects, anal atresia, cardiac defects, tracheoesophageal fistula, renal anomalies, limb abnormalities) with exposure to a TNF inhibitor,^25,84-86 but further reports have alleviated these concerns. Active transplacental transport occurs for adalimumab, infliximab, and golimumab,⁸⁷ but given structural differences, transport of certolizumab and etanercept is substantially less.^88,89 In the CRIB study of placental transfer of certolizumab from mother to infant (N=14), pharmacokinetic data demonstrated no quantifiable certolizumab levels in 13 infants and minimal levels in 1 infant at birth.⁸⁸ There are fewer data available on the use of other biologics in pregnancy, but for those in which active placental transport is relevant, similar concerns (ie, immunosuppression) might arise (eTable 3).

Concern over biologics largely involves risk for newborn immunosuppression. A case report detailed a Crohn disease patient treated with infliximab who gave birth to an infant who died of disseminated bacille Calmette-Guérin infection at 4.5 months after receiving the vaccine at 3 months.⁹⁰ This case underscores the importance of delaying live vaccination in infants born to mothers who were treated with a biologic during pregnancy. Authors have provided various data on how long to avoid vaccination; some state as long as 1 year.⁹¹

In pregnant females with inflammatory bowel disease treated with a biologic, no correlation was observed among maternal, placental, and infant serum biologic levels and neonatal infection. However, an association between preterm birth and the level of the biologic in maternal and placental (but not infant) serum and preterm birth was observed.⁹²

In another report from the same registry, combination therapy with a TNF inhibitor and another immunomodulator led to an increased risk for infection in infants at 12 months of age, compared to infants exposed to monotherapy⁸⁹ or exposed to neither agent.⁹³ A strategy to circumvent this potential problem is to avoid treatment with actively transported molecules in the third trimester.

Conclusion

Limited data exist to guide providers who are treating pregnant women with psoriasis. Our understanding of treatment of psoriasis in pregnancy is limited as a consequence of regulations surrounding clinical trials and inadequate detection of pregnancies in registries. Further efforts are necessary to better understand the relationship between psoriasis and pregnancy and how to manage pregnant women with psoriasis.

Historically, there have been limited data available on the management of psoriasis in pregnancy. The most comprehensive discussion of treatment guidelines is from 2012.¹ In the interim, many biologics have been approved for treating psoriasis, with slow accumulation of pregnancy safety data. The 2019 American Academy of Dermatology–National Psoriasis Foundation guidelines on biologics for psoriasis contain updated information but also highlight the paucity of pregnancy safety data.² This gap is in part a consequence of the exclusion and disenrollment of pregnant women from clinical trials.³ Additionally, lack of detection through registries contributes; pregnancy capture in registries is low compared to the expected number of pregnancies estimated from US Census data.⁴ Despite these shortcomings, psoriasis patients who are already pregnant or are considering becoming pregnant frequently are encountered in practice and may need treatment. This article reviews the evidence on commonly used treatments for psoriasis in pregnancy.

Background

For many patients, psoriasis improves during pregnancy^5,6 and becomes worse postpartum. In a prospective study, most patients reported improvement in pregnancy corresponding to a significant decrease in affected body surface area (P<.001) by 10 to 20 weeks’ gestation. Most patients also reported worsening of psoriasis postpartum; a significant increase in psoriatic body surface area (P=.001) was observed after delivery.⁷ Despite these findings, a considerable number of patients also experience stable disease or worsening of disease during pregnancy.

In addition to the maternal disease state, the issue of pregnancy outcomes is paramount. In the inflammatory bowel disease and rheumatology literature, it is established that uncontrolled disease is associated with poorer pregnancy outcomes.^8-10 Guidelines vary among societies on the use of biologics in pregnancy generally (eTable 1^1,2,9,11-24), but some societies recommend systemic agents to achieve disease control during pregnancy.^9,25

Assessing the potential interplay between disease severity and outcomes in pregnant women with psoriasis is further complicated by the slowly growing body of literature demonstrating that women with psoriasis have more comorbidities²⁶ and worse pregnancy outcomes.^27,28 Pregnant psoriasis patients are more likely to smoke, have depression, and be overweight or obese prior to pregnancy and are less likely to take prenatal vitamins.²⁶ They also have an increased risk for cesarean birth, gestational diabetes, gestational hypertension, and preeclampsia.²⁸ In contrast to these prior studies, a systematic review revealed no risk for adverse outcomes in pregnant women with psoriasis.²⁹

Assessment of Treatments for Psoriasis in Pregnancy

In light of these issues, treatment of psoriasis during pregnancy should be assessed from several vantage points. Of note, the US Food and Drug Administration changed its classification scheme in 2015 to a more narrative format called the Pregnancy and Lactation Labeling Rule.³⁰ Prior classifications, however, provide a reasonable starting point for categorizing the safety of drugs (Table³¹). Importantly, time of exposure to systemic agents also matters; first-trimester exposure is more likely to affect embryogenesis, whereas second- and third-trimester exposures are more prone to affect other aspects of fetal growth. eTable 2 provides data on the use of oral and topical medications to treat psoriasis in pregnancy.^1,8,22,32-45

Topical Agents
Topical steroids are largely understood to be reasonable treatment options, though consideration of potency, formulation, area of application, and use of occlusion is important.^1,46 Risk for orofacial cleft has been noted with first-trimester topical steroid exposure, though a 2015 Cochrane review update determined that the relative risk of this association was not significantly elevated.³²

The impact of topical calcipotriene and salicylic acid has not been studied in human pregnancies,¹ but systemic absorption can occur for both. There is potential for vitamin D toxicity with calcipotriene⁴⁶; consequently, use during pregnancy is not recommended.^1,46 Some authors recommend against topical salicylic acid in pregnancy; others report that limited exposure is permissible.⁴⁷ In fact, as salicylic acid commonly is found in over-the-counter acne products, many women of childbearing potential likely have quotidian exposure.

Preterm delivery and low birthweight have been reported with oral tacrolimus; however, risk with topical tacrolimus is thought to be low¹ because the molecular size likely prohibits notable absorption.⁴⁷ Evidence for the use of anthralin and coal tar also is scarce. First-trimester coal tar use should be avoided; subsequent use in pregnancy should be restricted given concern for adverse outcomes.¹

Phototherapy
Broadband or narrowband UVB therapy is recommended as second-line therapy in pregnancy. No cases of fetal risk or premature delivery associated with UVB therapy were found in our search.¹ Phototherapy can exacerbate melasma⁴⁷ and decrease folate levels⁴⁸; as such, some authors recommend folate supplementation in females of childbearing age who are being treated with phototherapy.⁴⁹ Psoralen, used in psoralen plus UVA therapy, is mutagenic and therefore contraindicated in pregnancy.¹

Oral Medications
Both methotrexate, which is a teratogen, abortifacient, and mutagen,¹ and systemic retinoids, which are teratogens, are contraindicated in pregnancy.^1,47 Acitretin labeling recommends avoiding pregnancy for 3 years posttreatment⁵⁰ because alcohol intake prolongs the medication’s half-life.²²

Apremilast use is not documented in pregnant psoriasis patients⁵¹; an ongoing registry of the Organization of Tetralogy Information Specialists has not reported publicly to date.⁵² Animal studies of apremilast have documented dose-related decreased birthweight and fetal loss.²²

Safety data for systemic steroids, used infrequently in psoriasis, are not well established. First-trimester prednisone exposure has been associated with prematurity, low birthweight, and congenital abnormalities.³⁸ A separate evaluation of 1047 children exposed to betamethasone in utero failed to demonstrate significant change in birthweight or head circumference. However, repeat antenatal corticosteroid exposure was associated with attention problems at 2 years of age.³⁹

Data regarding cyclosporine use, derived primarily from organ transplant recipients, suggest elevated risk for prematurity and low birthweight.^53,54 A meta-analysis demonstrated that organ transplant recipients taking cyclosporine had a nonsignificantly elevated odds ratio for congenital malformations, prematurity, and low birthweight.⁴² Cyclosporine use for psoriasis in pregnancy is not well described; in a study, rates of prematurity and low birthweight were both 21%.⁴³ Limited data are available for Janus kinase inhibitors, none of which are approved for psoriasis, though clinical trials in psoriasis and psoriatic arthritis are underway (ClinicalTrials.gov identifiers NCT04246372, NCT03104374, NCT03104400).

Biologics and Small-Molecule Inhibitors
Limited data on biologics in pregnancy exist²⁵ (eTable 3). Placental transport of IgG antibodies, including biologics, increases throughout pregnancy, especially in the third trimester.⁸² Infants of mothers treated with a biologic with potential for placental transfer are therefore considered by some authors to be immunosuppressed during the first months of life.²

Looking globally across biologics used for psoriasis, limited safety data are encouraging. In a review of PSOLAR (Psoriasis Longitudinal Assessment and Registry), 83 pregnancies with biologic exposure resulted in 59 live births (71%); 18 spontaneous abortions (22%); 6 induced abortions (7%); no congenital abnormalities; and 7 reports of neonatal problems, including respiratory issues, ABO blood group mismatch, hospitalization, and opioid withdrawal.⁸³

Use of tumor necrosis factor (TNF) inhibitors in pregnancy has the most data²⁵ and is considered a reasonable treatment option. Historically, there was concern about the risk for VACTERL syndrome (vertebral defects, anal atresia, cardiac defects, tracheoesophageal fistula, renal anomalies, limb abnormalities) with exposure to a TNF inhibitor,^25,84-86 but further reports have alleviated these concerns. Active transplacental transport occurs for adalimumab, infliximab, and golimumab,⁸⁷ but given structural differences, transport of certolizumab and etanercept is substantially less.^88,89 In the CRIB study of placental transfer of certolizumab from mother to infant (N=14), pharmacokinetic data demonstrated no quantifiable certolizumab levels in 13 infants and minimal levels in 1 infant at birth.⁸⁸ There are fewer data available on the use of other biologics in pregnancy, but for those in which active placental transport is relevant, similar concerns (ie, immunosuppression) might arise (eTable 3).

Concern over biologics largely involves risk for newborn immunosuppression. A case report detailed a Crohn disease patient treated with infliximab who gave birth to an infant who died of disseminated bacille Calmette-Guérin infection at 4.5 months after receiving the vaccine at 3 months.⁹⁰ This case underscores the importance of delaying live vaccination in infants born to mothers who were treated with a biologic during pregnancy. Authors have provided various data on how long to avoid vaccination; some state as long as 1 year.⁹¹

In pregnant females with inflammatory bowel disease treated with a biologic, no correlation was observed among maternal, placental, and infant serum biologic levels and neonatal infection. However, an association between preterm birth and the level of the biologic in maternal and placental (but not infant) serum and preterm birth was observed.⁹²

In another report from the same registry, combination therapy with a TNF inhibitor and another immunomodulator led to an increased risk for infection in infants at 12 months of age, compared to infants exposed to monotherapy⁸⁹ or exposed to neither agent.⁹³ A strategy to circumvent this potential problem is to avoid treatment with actively transported molecules in the third trimester.

Conclusion

Limited data exist to guide providers who are treating pregnant women with psoriasis. Our understanding of treatment of psoriasis in pregnancy is limited as a consequence of regulations surrounding clinical trials and inadequate detection of pregnancies in registries. Further efforts are necessary to better understand the relationship between psoriasis and pregnancy and how to manage pregnant women with psoriasis.

The use of lasers in dermatology has evolved and expanded since their first cutaneous use in 1963.¹ As the fundamental understanding of the interaction of laser energy with biological tissues increased, the need for laser safety became apparent. Since then, lasers of varying wavelengths have been developed, each with its specific chromophore target and specific safety need. Protocols, such as a checklist, that have been shown to reduce adverse events in surgery and in the intensive care unit can be borrowed to decrease risk from laser injury and optimize laser safety in dermatology.² The safety of the patient, the laser operator, and the other health care providers involved in the delivery of laser therapy led to the first US Food and Drug Administration (FDA) guidelines for laser use in 1984.³

There are 4 regulatory organizations for laser safety in the United States: the American National Standards Institute (ANSI), the Occupational Health and Safety Administration (OSHA), the FDA’s Center for Devices and Radiological Health, and The Joint Commission. The American National Standards Institute is a nonprofit group composed of laser manufacturers, government agencies, professional societies, educational institutions, and consumer and labor groups. It publishes voluntary safety standards and periodic updates (the series is labelled ANSI Z136) for the use of lasers in general (ANSI Z136.1) and for health care use in particular (ANSI Z136.3), including their use in dermatology. Laser hazard classifications also originate from ANSI. The standards of care established by ANSI guidelines are those by which health care providers are judged in health care litigation and are used by the other 3 organizations listed above. The Center for Devices and Radiological Health oversees laser manufacturers and their adherence to safety standards, determines laser hazard classifications such as ANSI, and requires manufacturers to affix a hazard class to the laser when manufactured. The Joint Commission is the accreditation body for health care programs and inspects hospitals and clinics for compliance with ANSI standards. Additionally, the American Society for Laser Medicine and Surgery, the American Academy of Dermatology, and the American Society for Dermatologic Surgery are professional organizations involved in laser operational safety training.³

Laser Principles

The basic principles of lasers include transmission, absorption, scatter, and reflection, all occurring when laser light is applied to biological tissues. The effects of the laser are a function of the target tissue (the chromophore) and the wavelength of light being used.⁴ In the skin, there are 3 main endogenous chromophores: water, hemoglobin, and melanin. Some experts consider collagen to be a fourth and separate entity as a chromophore. Tattoos are considered exogenous chromophores.³ The basic principles of lasers are important to understand and keep in mind when discussing laser safety, as they are the mechanisms through which unintended consequences can occur.

Laser Safety

Ocular Hazards
Ocular hazards are a notable concern in laser surgery. The eye is uniquely susceptible to laser light, and eye injuries represent a majority of reported injuries, which can occur through direct beam, mirror reflection by surgical instruments, and beam reflection off the skin (4%–7% of light that hits the skin is reflected because of the refractive index between air and the stratum corneum).³ The different wavelengths of lasers affect different parts of the eye. The 3 parts of the eye affected most are the retina, cornea, and lens. Not only is the lens primarily at risk for acute (lenticular burns) and chronic (cataracts) injury from the laser, but secondarily the lens also can concentrate a laser beam onto the retina by a factor of 100,000 (Table 1).³

The use of ocular protective equipment, sometimes referred to as personal protective eyewear (PPE), is essential and is mandated by ANSI and OSHA for all class 3 and class 4 lasers. The eyewear must be labeled with the wavelength and the degree of optical protection—termed the optical density (OD) or filter factor—of each lens and should match the laser being used. Laser manufacturers, as required by ANSI, must provide the wavelength and OD of their lasers, and both can be found on each laser as well as in ANSI Z136.1.³

Vendors supplying PPE generally provide the material, usually glass or polycarbonate; color; visible light transmission, which is the actual amount of light that reaches one’s eye through the lens; filter specifications, which contain the OD at certain wavelengths; and the types of lasers for which each specific PPE is used. It is important to match the laser to the correct PPE. The use of multiple types of lasers in the same office or laser treatment area can present challenges regarding eye safety. Matching the PPE to the laser in use is critical, and therefore all steps to prevent error for patients and personnel should be employed. One recommendation is to place each laser in a separate room with the appropriate PPE hung outside on the door of that room.

When the treatment area is in the periocular region, protection of the patient’s cornea is essential. Leaded eye shields with nonreflective surfaces have been shown to offer the best protection.⁵ Prior to placement, anesthetic eye drops and lubrication are important for patient comfort and protection from corneal injury.

Laser-Generated Airborne Contaminants
Other hazards associated with laser use not directly related to the beam are laser-generated airborne contaminants (LGACs), including chemicals, viruses, bacteria, aerosolized blood products, and nanoparticles (<1 µm) known as ultrafine particles (UFPs). According to ANSI, electrosurgical devices and lasers generate the same smoke. The plume (surgical smoke) is known to contain as many as 60 chemicals, including but not limited to carbon monoxide, acrylonitrite, hydrocyanide, benzene, toluene, naphthalene, and formaldehyde. Several are known carcinogens, and others are environmental toxins.^6,7

Smoke management is an important consideration for dermatologists and their patients and generally includes respiratory protection via masks and ventilation techniques. However, the practice is not universal, and oversight agencies such as OSHA and the National Institute for Occupational Safety and Health (NIOSH) provide guidelines only; they do not enforce. As such, smoke management is voluntary and not widely practiced. In a 2014 survey of 997 dermatologic surgeons who were asked if smoke management is used in their practice, 77% of respondents indicated no smoke management was used.⁶

The Surgical Plume: Composition
A 2014 study from the University of California, San Diego Department of Dermatology analyzed surgical smoke.⁶ The researchers placed the smoke collection probe 16 to 18 inches above the electrocautery site, which approximates the location of the surgeon’s head during the procedure. Assessing smoke composition, they found high levels of carcinogens and irritants. Two compounds found in their assay—1,3-butadiene and benzene—also are found in secondhand cigarette smoke. However, the concentrations in the plume were 17-fold higher for 1,3-butadiene and 10-fold higher for benzene than those found in secondhand cigarette smoke. The risk from chronic, long-term exposure to these airborne contaminants is notable, as benzene (a known carcinogen as determined by the US Department of Health and Human Services) is known to cause leukemia. For example, a busy Mohs surgeon can reach the equivalent of as many as 50 hours of continuous smoke exposure over the course of a year.⁶

The Surgical Plume: Particle Concentration
Ultrafine particles can bypass conventional filtering systems (surgical masks and N95 respirators) because of their extremely small size, which allows them to pass further into the lungs and all the way to the alveolar spaces. Geographic regions with high UFPs have been shown to have higher overall mortality rates, as well as higher rates of reactive airway disease, cardiovascular disease, and lung cancer. A 2016 study by Chuang et al⁷ published in JAMA Dermatology looked at the UFPs in the surgical plume from laser hair removal (LHR) procedures. The plume of LHR has a distinct odor and easily discernible particulates. The investigators measured the UFPs at the level of the laser practitioner and the patient’s face during LHR with a smoke evacuator turned on and again with it turned off for 30 seconds, and then compared them to UFPs measured in the treatment room, the waiting room, and outside the building. There were substantial increases in UFPs from the LHR procedure, especially for the laser practitioner, when the smoke evacuator was off. The ambient baseline particle count, as measured in the clinic waiting area, began at 15,300 particles per cubic centimeter (PPC), and once the LHR procedure began (smoke evacuator on), there was a greater than 8-fold PPC increase above baseline (15,300 PPC to 129,376 PPC) in UFPs measured for the laser practitioner. Importantly, during LHR when the smoke evacuator was turned off for 30 seconds, there was a more than 28-fold increase (15,300 PPC to 435,888 PPC) over baseline to the practitioner (Figure).⁷

The Surgical Plume: Smoke Management
Many virus particles and UFPs are less than 0.1 µm in size. It is important to note that neither surgical masks nor high-filtration masks, such as the N95 respirator, filter particles smaller than 0.1 µm. The first line of defense in smoke management is the local exhaust ventilation (LEV) system, which includes wall suction and/or a smoke evacuator. The smoke evacuator is considered the more important of the two. General filtration, such as wall suction, is a low-flow system and is really used for liquids. It can be used as a supplement to the smoke evacuator to control small amounts of plume if fitted with an in-line filter. There are 2 types of LEV filters: ultralow particulate air filters filter particles larger than 0.1µm, whereas high-efficiency particulate air filters filter particles larger than 0.3 µm. The ultralow particulate filters are used in most of the newer LEVs in use today and filter 0.1-µm particles at 99.99% efficiency.³

Of utmost importance when using a smoke evacuator system is suction tip placement. Placing the suction tip 1 cm from the tissue damage site has been shown to be 98.6% effective at removing laser plume. If moved to 2 cm, effectiveness decreases to less than 50%.¹¹ Proper management recommendations based on current evidence suggest that use of a smoke evacuator and an approved fit-tested N95 respirator might provide maximum protection.⁶ In addition to plume exposure, tissue splatter can occur, especially during ablative (CO₂) and tattoo laser therapy, which should prompt consideration of a face shield.¹¹ There are several vendors and models available online, and a simple Internet search for surgical tissue splatter face shields will provide multiple options.

The standard surgical mask is not NIOSH approved and only effectively (99%) filters particles larger than 5 µm (vs 25% efficacy for 0.3-µm particles). Its main purpose is to protect the patient from the wearer.¹²

High-filtration masks, which capture particles as small as 0.1 µm, should be used instead. The surgical N95 respirator is a NIOSH-certified respirator and is recommended for use in cases when smoke management is necessary. The FDA does not test or certify these masks; it only clears them after reviewing manufacturer test data. Technically, to be called a surgical mask, it must be cleared by the FDA.¹² The 95 of N95 indicates filter efficiency ratings of 95% when testing the filter efficiency using particles of approximately 0.3 µm in diameter (Table 2).¹³ Because 77% of surgical smoke particles are smaller than 1.1 µm, surgical masks and N95 respirators are never sufficient as stand-alone protection.¹⁴ An LEV system is much more important for safe surgical smoke management. However, recommendations call for the use of a smoke evacuator and a high-filtration mask together to obtain the most protection available.¹⁴

Checklists

Checklists are ubiquitous throughout many occupations and many medical specialties. Their usefulness in preventing adverse events is well established. Any patient-provider encounter in which a series of sequential actions is required is a perfect situation for a checklist. In dermatologic laser surgery where the eye is uniquely susceptible to injury, a laser safety checklist is essential. Additionally, there are issues with LGACs and fire that are important to consider. Having protocols (ie, a checklist) in place that address these safety issues has been shown to reduce adverse outcomes.² There are a number of templates available from various sources that can be customized to the laser treatment area. We provide a modifiable example (Table 3).

Conclusion

Laser usage in dermatologic surgery has increased. According to surveys from the American Society for Dermatologic Surgery, in 2012 there were approximately 2 million laser/light/energy-based procedures performed. By 2017, there were 3.27 million, up from 2.79 million in 2016, representing an approximate 1-year increase of 17%.¹⁶ Lasers have allowed interventions for skin, vascular, and aesthetic conditions that were once untreatable. As their use increases in number and broadens in scope, there also has been an increase in litigation alleging malpractice for misuse of the laser.¹⁷ Adverse events, which include photochemical or thermal injuries to the skin, pigmentation issues, scarring, plume-related issues, and fires, do occur. One solution to reduce the chance of an adverse outcome is to implement a checklist. Research using checklists has shown that adverse events are reduced when checklists are created and implemented properly. Improving checklist compliance also improves patient outcomes.¹⁷ The American National Standards Institute, in their ANSI Z136 series, and the World Health Organization provide checklist templates. We include our checklist for use in laser surgery (Table 3). Understanding that each laser treatment area is unique, the templates can serve as a starting point and can then be customized to suit the needs of each dermatologist.

The use of lasers in dermatology has evolved and expanded since their first cutaneous use in 1963.¹ As the fundamental understanding of the interaction of laser energy with biological tissues increased, the need for laser safety became apparent. Since then, lasers of varying wavelengths have been developed, each with its specific chromophore target and specific safety need. Protocols, such as a checklist, that have been shown to reduce adverse events in surgery and in the intensive care unit can be borrowed to decrease risk from laser injury and optimize laser safety in dermatology.² The safety of the patient, the laser operator, and the other health care providers involved in the delivery of laser therapy led to the first US Food and Drug Administration (FDA) guidelines for laser use in 1984.³

There are 4 regulatory organizations for laser safety in the United States: the American National Standards Institute (ANSI), the Occupational Health and Safety Administration (OSHA), the FDA’s Center for Devices and Radiological Health, and The Joint Commission. The American National Standards Institute is a nonprofit group composed of laser manufacturers, government agencies, professional societies, educational institutions, and consumer and labor groups. It publishes voluntary safety standards and periodic updates (the series is labelled ANSI Z136) for the use of lasers in general (ANSI Z136.1) and for health care use in particular (ANSI Z136.3), including their use in dermatology. Laser hazard classifications also originate from ANSI. The standards of care established by ANSI guidelines are those by which health care providers are judged in health care litigation and are used by the other 3 organizations listed above. The Center for Devices and Radiological Health oversees laser manufacturers and their adherence to safety standards, determines laser hazard classifications such as ANSI, and requires manufacturers to affix a hazard class to the laser when manufactured. The Joint Commission is the accreditation body for health care programs and inspects hospitals and clinics for compliance with ANSI standards. Additionally, the American Society for Laser Medicine and Surgery, the American Academy of Dermatology, and the American Society for Dermatologic Surgery are professional organizations involved in laser operational safety training.³

Laser Principles

The basic principles of lasers include transmission, absorption, scatter, and reflection, all occurring when laser light is applied to biological tissues. The effects of the laser are a function of the target tissue (the chromophore) and the wavelength of light being used.⁴ In the skin, there are 3 main endogenous chromophores: water, hemoglobin, and melanin. Some experts consider collagen to be a fourth and separate entity as a chromophore. Tattoos are considered exogenous chromophores.³ The basic principles of lasers are important to understand and keep in mind when discussing laser safety, as they are the mechanisms through which unintended consequences can occur.

Laser Safety

Ocular Hazards
Ocular hazards are a notable concern in laser surgery. The eye is uniquely susceptible to laser light, and eye injuries represent a majority of reported injuries, which can occur through direct beam, mirror reflection by surgical instruments, and beam reflection off the skin (4%–7% of light that hits the skin is reflected because of the refractive index between air and the stratum corneum).³ The different wavelengths of lasers affect different parts of the eye. The 3 parts of the eye affected most are the retina, cornea, and lens. Not only is the lens primarily at risk for acute (lenticular burns) and chronic (cataracts) injury from the laser, but secondarily the lens also can concentrate a laser beam onto the retina by a factor of 100,000 (Table 1).³

The use of ocular protective equipment, sometimes referred to as personal protective eyewear (PPE), is essential and is mandated by ANSI and OSHA for all class 3 and class 4 lasers. The eyewear must be labeled with the wavelength and the degree of optical protection—termed the optical density (OD) or filter factor—of each lens and should match the laser being used. Laser manufacturers, as required by ANSI, must provide the wavelength and OD of their lasers, and both can be found on each laser as well as in ANSI Z136.1.³

Vendors supplying PPE generally provide the material, usually glass or polycarbonate; color; visible light transmission, which is the actual amount of light that reaches one’s eye through the lens; filter specifications, which contain the OD at certain wavelengths; and the types of lasers for which each specific PPE is used. It is important to match the laser to the correct PPE. The use of multiple types of lasers in the same office or laser treatment area can present challenges regarding eye safety. Matching the PPE to the laser in use is critical, and therefore all steps to prevent error for patients and personnel should be employed. One recommendation is to place each laser in a separate room with the appropriate PPE hung outside on the door of that room.

When the treatment area is in the periocular region, protection of the patient’s cornea is essential. Leaded eye shields with nonreflective surfaces have been shown to offer the best protection.⁵ Prior to placement, anesthetic eye drops and lubrication are important for patient comfort and protection from corneal injury.

Laser-Generated Airborne Contaminants
Other hazards associated with laser use not directly related to the beam are laser-generated airborne contaminants (LGACs), including chemicals, viruses, bacteria, aerosolized blood products, and nanoparticles (<1 µm) known as ultrafine particles (UFPs). According to ANSI, electrosurgical devices and lasers generate the same smoke. The plume (surgical smoke) is known to contain as many as 60 chemicals, including but not limited to carbon monoxide, acrylonitrite, hydrocyanide, benzene, toluene, naphthalene, and formaldehyde. Several are known carcinogens, and others are environmental toxins.^6,7

Smoke management is an important consideration for dermatologists and their patients and generally includes respiratory protection via masks and ventilation techniques. However, the practice is not universal, and oversight agencies such as OSHA and the National Institute for Occupational Safety and Health (NIOSH) provide guidelines only; they do not enforce. As such, smoke management is voluntary and not widely practiced. In a 2014 survey of 997 dermatologic surgeons who were asked if smoke management is used in their practice, 77% of respondents indicated no smoke management was used.⁶

The Surgical Plume: Composition
A 2014 study from the University of California, San Diego Department of Dermatology analyzed surgical smoke.⁶ The researchers placed the smoke collection probe 16 to 18 inches above the electrocautery site, which approximates the location of the surgeon’s head during the procedure. Assessing smoke composition, they found high levels of carcinogens and irritants. Two compounds found in their assay—1,3-butadiene and benzene—also are found in secondhand cigarette smoke. However, the concentrations in the plume were 17-fold higher for 1,3-butadiene and 10-fold higher for benzene than those found in secondhand cigarette smoke. The risk from chronic, long-term exposure to these airborne contaminants is notable, as benzene (a known carcinogen as determined by the US Department of Health and Human Services) is known to cause leukemia. For example, a busy Mohs surgeon can reach the equivalent of as many as 50 hours of continuous smoke exposure over the course of a year.⁶

The Surgical Plume: Particle Concentration
Ultrafine particles can bypass conventional filtering systems (surgical masks and N95 respirators) because of their extremely small size, which allows them to pass further into the lungs and all the way to the alveolar spaces. Geographic regions with high UFPs have been shown to have higher overall mortality rates, as well as higher rates of reactive airway disease, cardiovascular disease, and lung cancer. A 2016 study by Chuang et al⁷ published in JAMA Dermatology looked at the UFPs in the surgical plume from laser hair removal (LHR) procedures. The plume of LHR has a distinct odor and easily discernible particulates. The investigators measured the UFPs at the level of the laser practitioner and the patient’s face during LHR with a smoke evacuator turned on and again with it turned off for 30 seconds, and then compared them to UFPs measured in the treatment room, the waiting room, and outside the building. There were substantial increases in UFPs from the LHR procedure, especially for the laser practitioner, when the smoke evacuator was off. The ambient baseline particle count, as measured in the clinic waiting area, began at 15,300 particles per cubic centimeter (PPC), and once the LHR procedure began (smoke evacuator on), there was a greater than 8-fold PPC increase above baseline (15,300 PPC to 129,376 PPC) in UFPs measured for the laser practitioner. Importantly, during LHR when the smoke evacuator was turned off for 30 seconds, there was a more than 28-fold increase (15,300 PPC to 435,888 PPC) over baseline to the practitioner (Figure).⁷

The Surgical Plume: Smoke Management
Many virus particles and UFPs are less than 0.1 µm in size. It is important to note that neither surgical masks nor high-filtration masks, such as the N95 respirator, filter particles smaller than 0.1 µm. The first line of defense in smoke management is the local exhaust ventilation (LEV) system, which includes wall suction and/or a smoke evacuator. The smoke evacuator is considered the more important of the two. General filtration, such as wall suction, is a low-flow system and is really used for liquids. It can be used as a supplement to the smoke evacuator to control small amounts of plume if fitted with an in-line filter. There are 2 types of LEV filters: ultralow particulate air filters filter particles larger than 0.1µm, whereas high-efficiency particulate air filters filter particles larger than 0.3 µm. The ultralow particulate filters are used in most of the newer LEVs in use today and filter 0.1-µm particles at 99.99% efficiency.³

Of utmost importance when using a smoke evacuator system is suction tip placement. Placing the suction tip 1 cm from the tissue damage site has been shown to be 98.6% effective at removing laser plume. If moved to 2 cm, effectiveness decreases to less than 50%.¹¹ Proper management recommendations based on current evidence suggest that use of a smoke evacuator and an approved fit-tested N95 respirator might provide maximum protection.⁶ In addition to plume exposure, tissue splatter can occur, especially during ablative (CO₂) and tattoo laser therapy, which should prompt consideration of a face shield.¹¹ There are several vendors and models available online, and a simple Internet search for surgical tissue splatter face shields will provide multiple options.

The standard surgical mask is not NIOSH approved and only effectively (99%) filters particles larger than 5 µm (vs 25% efficacy for 0.3-µm particles). Its main purpose is to protect the patient from the wearer.¹²

High-filtration masks, which capture particles as small as 0.1 µm, should be used instead. The surgical N95 respirator is a NIOSH-certified respirator and is recommended for use in cases when smoke management is necessary. The FDA does not test or certify these masks; it only clears them after reviewing manufacturer test data. Technically, to be called a surgical mask, it must be cleared by the FDA.¹² The 95 of N95 indicates filter efficiency ratings of 95% when testing the filter efficiency using particles of approximately 0.3 µm in diameter (Table 2).¹³ Because 77% of surgical smoke particles are smaller than 1.1 µm, surgical masks and N95 respirators are never sufficient as stand-alone protection.¹⁴ An LEV system is much more important for safe surgical smoke management. However, recommendations call for the use of a smoke evacuator and a high-filtration mask together to obtain the most protection available.¹⁴

Checklists

Checklists are ubiquitous throughout many occupations and many medical specialties. Their usefulness in preventing adverse events is well established. Any patient-provider encounter in which a series of sequential actions is required is a perfect situation for a checklist. In dermatologic laser surgery where the eye is uniquely susceptible to injury, a laser safety checklist is essential. Additionally, there are issues with LGACs and fire that are important to consider. Having protocols (ie, a checklist) in place that address these safety issues has been shown to reduce adverse outcomes.² There are a number of templates available from various sources that can be customized to the laser treatment area. We provide a modifiable example (Table 3).

Conclusion

Laser usage in dermatologic surgery has increased. According to surveys from the American Society for Dermatologic Surgery, in 2012 there were approximately 2 million laser/light/energy-based procedures performed. By 2017, there were 3.27 million, up from 2.79 million in 2016, representing an approximate 1-year increase of 17%.¹⁶ Lasers have allowed interventions for skin, vascular, and aesthetic conditions that were once untreatable. As their use increases in number and broadens in scope, there also has been an increase in litigation alleging malpractice for misuse of the laser.¹⁷ Adverse events, which include photochemical or thermal injuries to the skin, pigmentation issues, scarring, plume-related issues, and fires, do occur. One solution to reduce the chance of an adverse outcome is to implement a checklist. Research using checklists has shown that adverse events are reduced when checklists are created and implemented properly. Improving checklist compliance also improves patient outcomes.¹⁷ The American National Standards Institute, in their ANSI Z136 series, and the World Health Organization provide checklist templates. We include our checklist for use in laser surgery (Table 3). Understanding that each laser treatment area is unique, the templates can serve as a starting point and can then be customized to suit the needs of each dermatologist.

The use of lasers in dermatology has evolved and expanded since their first cutaneous use in 1963.¹ As the fundamental understanding of the interaction of laser energy with biological tissues increased, the need for laser safety became apparent. Since then, lasers of varying wavelengths have been developed, each with its specific chromophore target and specific safety need. Protocols, such as a checklist, that have been shown to reduce adverse events in surgery and in the intensive care unit can be borrowed to decrease risk from laser injury and optimize laser safety in dermatology.² The safety of the patient, the laser operator, and the other health care providers involved in the delivery of laser therapy led to the first US Food and Drug Administration (FDA) guidelines for laser use in 1984.³

There are 4 regulatory organizations for laser safety in the United States: the American National Standards Institute (ANSI), the Occupational Health and Safety Administration (OSHA), the FDA’s Center for Devices and Radiological Health, and The Joint Commission. The American National Standards Institute is a nonprofit group composed of laser manufacturers, government agencies, professional societies, educational institutions, and consumer and labor groups. It publishes voluntary safety standards and periodic updates (the series is labelled ANSI Z136) for the use of lasers in general (ANSI Z136.1) and for health care use in particular (ANSI Z136.3), including their use in dermatology. Laser hazard classifications also originate from ANSI. The standards of care established by ANSI guidelines are those by which health care providers are judged in health care litigation and are used by the other 3 organizations listed above. The Center for Devices and Radiological Health oversees laser manufacturers and their adherence to safety standards, determines laser hazard classifications such as ANSI, and requires manufacturers to affix a hazard class to the laser when manufactured. The Joint Commission is the accreditation body for health care programs and inspects hospitals and clinics for compliance with ANSI standards. Additionally, the American Society for Laser Medicine and Surgery, the American Academy of Dermatology, and the American Society for Dermatologic Surgery are professional organizations involved in laser operational safety training.³

Laser Principles

The basic principles of lasers include transmission, absorption, scatter, and reflection, all occurring when laser light is applied to biological tissues. The effects of the laser are a function of the target tissue (the chromophore) and the wavelength of light being used.⁴ In the skin, there are 3 main endogenous chromophores: water, hemoglobin, and melanin. Some experts consider collagen to be a fourth and separate entity as a chromophore. Tattoos are considered exogenous chromophores.³ The basic principles of lasers are important to understand and keep in mind when discussing laser safety, as they are the mechanisms through which unintended consequences can occur.

Laser Safety

Ocular Hazards
Ocular hazards are a notable concern in laser surgery. The eye is uniquely susceptible to laser light, and eye injuries represent a majority of reported injuries, which can occur through direct beam, mirror reflection by surgical instruments, and beam reflection off the skin (4%–7% of light that hits the skin is reflected because of the refractive index between air and the stratum corneum).³ The different wavelengths of lasers affect different parts of the eye. The 3 parts of the eye affected most are the retina, cornea, and lens. Not only is the lens primarily at risk for acute (lenticular burns) and chronic (cataracts) injury from the laser, but secondarily the lens also can concentrate a laser beam onto the retina by a factor of 100,000 (Table 1).³

The use of ocular protective equipment, sometimes referred to as personal protective eyewear (PPE), is essential and is mandated by ANSI and OSHA for all class 3 and class 4 lasers. The eyewear must be labeled with the wavelength and the degree of optical protection—termed the optical density (OD) or filter factor—of each lens and should match the laser being used. Laser manufacturers, as required by ANSI, must provide the wavelength and OD of their lasers, and both can be found on each laser as well as in ANSI Z136.1.³

Vendors supplying PPE generally provide the material, usually glass or polycarbonate; color; visible light transmission, which is the actual amount of light that reaches one’s eye through the lens; filter specifications, which contain the OD at certain wavelengths; and the types of lasers for which each specific PPE is used. It is important to match the laser to the correct PPE. The use of multiple types of lasers in the same office or laser treatment area can present challenges regarding eye safety. Matching the PPE to the laser in use is critical, and therefore all steps to prevent error for patients and personnel should be employed. One recommendation is to place each laser in a separate room with the appropriate PPE hung outside on the door of that room.

When the treatment area is in the periocular region, protection of the patient’s cornea is essential. Leaded eye shields with nonreflective surfaces have been shown to offer the best protection.⁵ Prior to placement, anesthetic eye drops and lubrication are important for patient comfort and protection from corneal injury.

Laser-Generated Airborne Contaminants
Other hazards associated with laser use not directly related to the beam are laser-generated airborne contaminants (LGACs), including chemicals, viruses, bacteria, aerosolized blood products, and nanoparticles (<1 µm) known as ultrafine particles (UFPs). According to ANSI, electrosurgical devices and lasers generate the same smoke. The plume (surgical smoke) is known to contain as many as 60 chemicals, including but not limited to carbon monoxide, acrylonitrite, hydrocyanide, benzene, toluene, naphthalene, and formaldehyde. Several are known carcinogens, and others are environmental toxins.^6,7

Smoke management is an important consideration for dermatologists and their patients and generally includes respiratory protection via masks and ventilation techniques. However, the practice is not universal, and oversight agencies such as OSHA and the National Institute for Occupational Safety and Health (NIOSH) provide guidelines only; they do not enforce. As such, smoke management is voluntary and not widely practiced. In a 2014 survey of 997 dermatologic surgeons who were asked if smoke management is used in their practice, 77% of respondents indicated no smoke management was used.⁶

The Surgical Plume: Composition
A 2014 study from the University of California, San Diego Department of Dermatology analyzed surgical smoke.⁶ The researchers placed the smoke collection probe 16 to 18 inches above the electrocautery site, which approximates the location of the surgeon’s head during the procedure. Assessing smoke composition, they found high levels of carcinogens and irritants. Two compounds found in their assay—1,3-butadiene and benzene—also are found in secondhand cigarette smoke. However, the concentrations in the plume were 17-fold higher for 1,3-butadiene and 10-fold higher for benzene than those found in secondhand cigarette smoke. The risk from chronic, long-term exposure to these airborne contaminants is notable, as benzene (a known carcinogen as determined by the US Department of Health and Human Services) is known to cause leukemia. For example, a busy Mohs surgeon can reach the equivalent of as many as 50 hours of continuous smoke exposure over the course of a year.⁶

The Surgical Plume: Particle Concentration
Ultrafine particles can bypass conventional filtering systems (surgical masks and N95 respirators) because of their extremely small size, which allows them to pass further into the lungs and all the way to the alveolar spaces. Geographic regions with high UFPs have been shown to have higher overall mortality rates, as well as higher rates of reactive airway disease, cardiovascular disease, and lung cancer. A 2016 study by Chuang et al⁷ published in JAMA Dermatology looked at the UFPs in the surgical plume from laser hair removal (LHR) procedures. The plume of LHR has a distinct odor and easily discernible particulates. The investigators measured the UFPs at the level of the laser practitioner and the patient’s face during LHR with a smoke evacuator turned on and again with it turned off for 30 seconds, and then compared them to UFPs measured in the treatment room, the waiting room, and outside the building. There were substantial increases in UFPs from the LHR procedure, especially for the laser practitioner, when the smoke evacuator was off. The ambient baseline particle count, as measured in the clinic waiting area, began at 15,300 particles per cubic centimeter (PPC), and once the LHR procedure began (smoke evacuator on), there was a greater than 8-fold PPC increase above baseline (15,300 PPC to 129,376 PPC) in UFPs measured for the laser practitioner. Importantly, during LHR when the smoke evacuator was turned off for 30 seconds, there was a more than 28-fold increase (15,300 PPC to 435,888 PPC) over baseline to the practitioner (Figure).⁷

The Surgical Plume: Smoke Management
Many virus particles and UFPs are less than 0.1 µm in size. It is important to note that neither surgical masks nor high-filtration masks, such as the N95 respirator, filter particles smaller than 0.1 µm. The first line of defense in smoke management is the local exhaust ventilation (LEV) system, which includes wall suction and/or a smoke evacuator. The smoke evacuator is considered the more important of the two. General filtration, such as wall suction, is a low-flow system and is really used for liquids. It can be used as a supplement to the smoke evacuator to control small amounts of plume if fitted with an in-line filter. There are 2 types of LEV filters: ultralow particulate air filters filter particles larger than 0.1µm, whereas high-efficiency particulate air filters filter particles larger than 0.3 µm. The ultralow particulate filters are used in most of the newer LEVs in use today and filter 0.1-µm particles at 99.99% efficiency.³

Of utmost importance when using a smoke evacuator system is suction tip placement. Placing the suction tip 1 cm from the tissue damage site has been shown to be 98.6% effective at removing laser plume. If moved to 2 cm, effectiveness decreases to less than 50%.¹¹ Proper management recommendations based on current evidence suggest that use of a smoke evacuator and an approved fit-tested N95 respirator might provide maximum protection.⁶ In addition to plume exposure, tissue splatter can occur, especially during ablative (CO₂) and tattoo laser therapy, which should prompt consideration of a face shield.¹¹ There are several vendors and models available online, and a simple Internet search for surgical tissue splatter face shields will provide multiple options.

The standard surgical mask is not NIOSH approved and only effectively (99%) filters particles larger than 5 µm (vs 25% efficacy for 0.3-µm particles). Its main purpose is to protect the patient from the wearer.¹²

High-filtration masks, which capture particles as small as 0.1 µm, should be used instead. The surgical N95 respirator is a NIOSH-certified respirator and is recommended for use in cases when smoke management is necessary. The FDA does not test or certify these masks; it only clears them after reviewing manufacturer test data. Technically, to be called a surgical mask, it must be cleared by the FDA.¹² The 95 of N95 indicates filter efficiency ratings of 95% when testing the filter efficiency using particles of approximately 0.3 µm in diameter (Table 2).¹³ Because 77% of surgical smoke particles are smaller than 1.1 µm, surgical masks and N95 respirators are never sufficient as stand-alone protection.¹⁴ An LEV system is much more important for safe surgical smoke management. However, recommendations call for the use of a smoke evacuator and a high-filtration mask together to obtain the most protection available.¹⁴

Checklists

Checklists are ubiquitous throughout many occupations and many medical specialties. Their usefulness in preventing adverse events is well established. Any patient-provider encounter in which a series of sequential actions is required is a perfect situation for a checklist. In dermatologic laser surgery where the eye is uniquely susceptible to injury, a laser safety checklist is essential. Additionally, there are issues with LGACs and fire that are important to consider. Having protocols (ie, a checklist) in place that address these safety issues has been shown to reduce adverse outcomes.² There are a number of templates available from various sources that can be customized to the laser treatment area. We provide a modifiable example (Table 3).

Conclusion

Laser usage in dermatologic surgery has increased. According to surveys from the American Society for Dermatologic Surgery, in 2012 there were approximately 2 million laser/light/energy-based procedures performed. By 2017, there were 3.27 million, up from 2.79 million in 2016, representing an approximate 1-year increase of 17%.¹⁶ Lasers have allowed interventions for skin, vascular, and aesthetic conditions that were once untreatable. As their use increases in number and broadens in scope, there also has been an increase in litigation alleging malpractice for misuse of the laser.¹⁷ Adverse events, which include photochemical or thermal injuries to the skin, pigmentation issues, scarring, plume-related issues, and fires, do occur. One solution to reduce the chance of an adverse outcome is to implement a checklist. Research using checklists has shown that adverse events are reduced when checklists are created and implemented properly. Improving checklist compliance also improves patient outcomes.¹⁷ The American National Standards Institute, in their ANSI Z136 series, and the World Health Organization provide checklist templates. We include our checklist for use in laser surgery (Table 3). Understanding that each laser treatment area is unique, the templates can serve as a starting point and can then be customized to suit the needs of each dermatologist.

Psoriasis is a systemic immune-mediated disorder characterized by erythematous, scaly, well-demarcated plaques on the skin that affects approximately 3% of the world’s population.¹ Although topical therapies often are the first-line treatment of mild to moderate psoriasis, approximately 1 in 6 individuals has moderate to severe disease that requires systemic treatment such as biologics or phototherapy.² In patients with localized disease that is refractory to treatment or who have moderate to severe psoriasis requiring systemic treatment, phototherapy should be considered as a potential low-risk treatment option.

In July 2019, the American Academy of Dermatology (AAD) and National Psoriasis Foundation (NPF) released an updated set of guidelines for the use of phototherapy in treating adult patients with psoriasis.³ Since the prior guidelines were released in 2010, there have been numerous studies affirming the efficacy of phototherapy, with several large meta-analyses helping to refine clinical recommendations.^4,5 Each treatment was ranked using Strength of Recommendation Taxonomy, with a score of A, B, or C based on the strength of the evidence supporting the given modality. With the ever-increasing number of treatment options for patients with psoriasis, these guidelines inform dermatologists of the recommendations for the initiation, maintenance, and optimization of phototherapy in the treatment of psoriasis.

The AAD-NPF recommendations discuss the mechanism of action, efficacy, safety, and frequency of adverse events of 10 commonly used phototherapy/photochemotherapy modalities. They also address dosing regimens, the potential to combine phototherapy with other therapies, and the efficacy of treatment modalities for different types of psoriasis.³ The purpose of this discussion is to present these guidelines in a condensed form for prescribers of phototherapy and to review the most clinically significant considerations during each step of treatment. Of note, we only highlight the treatment of adult patients and do not discuss information relevant to pediatric patients with psoriasis.

Choosing a Phototherapy Modality

Phototherapy may be considered for patients with psoriasis that affects more than 3% body surface area or for localized disease refractory to conventional treatments. UV light is believed to provide relief from psoriasis via multiple mechanisms, such as through favorable alterations in cytokine profiles, initiation of apoptosis, and local immunosupression.⁶ There is no single first-line phototherapeutic modality recommended for all patients with psoriasis. Rather, the decision to implement a particular modality should be individualized to the patient, considering factors such as percentage of body surface area affected by disease, quality-of-life assessment, comorbidities, lifestyle, and cost of treatment.

Of the 10 phototherapy modalities reviewed in these guidelines, 4 were ranked by the AAD and NPF as having grade A evidence for efficacy in the treatment of moderate to severe plaque psoriasis. Treatments with a grade A level of recommendation included narrowband UVB (NB-UVB), broadband UVB (BB-UVB), targeted phototherapy (excimer laser and excimer lamp), and oral psoralen plus UVA (PUVA) therapy. Photodynamic therapy for psoriasis was given an A-level recommendation against its use, as it was found to be ineffective with an unfavorable side-effect profile. Treatments with a grade B level of recommendation—nonoral routes of PUVA therapy, pulsed dye laser/intense pulsed light for nail psoriasis only, Goeckerman therapy, and climatotherapy—have sufficient evidence available to support their treatment of moderate to severe psoriasis in some cases. Treatments with a grade C level of recommendation—Grenz ray therapy (also called borderline or ultrasoft therapy) and visible light therapy—have insufficient evidence to support their use in patients with moderate to severe psoriasis (Table 1).

Psoralen plus UVA, which may be used topically (ie, bathwater PUVA) or taken orally, refers to treatment with photosensitizing psoralens. Psoralens are agents that intercalate with DNA and enhance the efficacy of phototherapy.¹⁰ Topical PUVA, with a grade B level of recommendation, is an effective treatment option for patients with localized disease and has been shown to be particularly efficacious in the treatment of palmoplantar pustular psoriasis. Oral PUVA is an effective option for psoriasis with a grade A recommendation, while bathwater PUVA has a grade B level of recommendation for moderate to severe plaque psoriasis. Oral PUVA is associated with greater systemic side effects (both acute and subacute) compared with NB-UVB and also is associated with photocarcinogenesis, particularly squamous cell carcinoma in white patients.¹¹ Other side effects from PUVA include pigmented macules in sun-protected areas (known as PUVA lentigines), which may make evaluation of skin lesions challenging. Because of the increased risk for cancer with oral PUVA, NB-UVB is preferable as a first-line treatment vs PUVA, especially in patients with a history of skin cancer.^12,13

Goeckerman therapy, which involves the synergistic combination of UVB and crude coal tar, is an older treatment that has shown efficacy in the treatment of severe or recalcitrant psoriasis (grade B level of recommendation). One prior case-control study comparing the efficacy of Goeckerman therapy with newer treatments, such as biologic therapies, steroids, and oral immunosuppressants, found a similar reduction in symptoms among both treatment groups, with longer disease-free periods in patients who received Goeckerman therapy than those who received newer therapies (22.3 years vs 4.6 months).¹⁴ However, Goeckerman therapy is utilized less frequently than more modern therapies because of the time required for treatment and declining insurance reimbursements for it. Climatotherapy, another older established therapy, involves the temporary or permanent relocation of patients to an environment that is favorable for disease control (grade B level of recommendation). Locations such as the Dead Sea and Canary Islands have been studied and shown to provide both subjective and objective improvement in patients’ psoriasis disease course. Patients had notable improvement in both their psoriasis area and severity index score and quality of life after a 3- to 4-week relocation to these areas.^15,16 Access to climatotherapy and the transient nature of disease relief are apparent limitations of this treatment modality.

Grenz ray is a type of phototherapy that uses longer-wavelength ionizing radiation, which has low penetrance into surrounding tissues and a 95% absorption rate within the first 3 mm of the skin depth. This treatment has been used less frequently since the development of newer alternatives but should still be considered as a second line to UV therapy, especially in cases of recalcitrant disease and palmoplantar psoriasis, and when access to other treatment options is limited. Grenz ray has a grade C level of recommendation due to the paucity of evidence that supports its efficacy. Thus, it is not recommended as first-line therapy for the treatment of moderate to severe psoriasis. Visible light therapy is another treatment option that uses light in the visible wavelength spectrum but predominantly utilizes blue and red light. Psoriatic lesions are sensitive to light therapy because of the elevated levels of naturally occurring photosensitizing agents, called protoporphyrins, in these lesions.¹⁷ Several small studies have shown improvement in psoriatic lesions treated with visible light therapy, with blue light showing greater efficacy in lesional clearance than red light.^18,19

Pulsed dye laser is a phototherapy modality that has shown efficacy in the treatment of nail psoriasis (grade B level of recommendation). One study comparing the effects of tazarotene cream 0.1% with pulsed dye laser and tazarotene cream 0.1% alone showed that patients receiving combination therapy had a greater decrease in nail psoriasis severity index scores, higher scores on the patient’s global assessment of improvement, and higher rates of improvement on the physician global assessment score. A physician global assessment score of 75% improvement or more was seen in patients treated with combination therapy vs monotherapy (5.3% vs 31.6%).²⁰ Intense pulsed light, a type of visible light therapy, also has been used to treat nail psoriasis, with one study showing notable improvement in nail bed and matrix disease and a global improvement in nail psoriasis severity index score after 6 months of biweekly treatment.²¹ However, this treatment has a grade B level of recommendation given the limited number of studies supporting the efficacy of this modality.

Initiation of Phototherapy

Prior to initiating phototherapy, it is important to assess the patient for any personal or family history of skin cancer, as phototherapy carries an increased risk for cutaneous malignancy, especially in patients with a history of skin cancer.^22,23 All patients also should be evaluated for their Fitzpatrick skin type, and the minimal erythema dose should be defined for those initiating UVB treatment. These classifications can be useful for the initial determination of treatment dose and the prevention of treatment-related adverse events (TRAEs). A careful drug history also should be taken before the initiation of phototherapy to avoid photosensitizing reactions. Thiazide diuretics and tetracyclines are 2 such examples of medications commonly associated with photosensitizing reactions.²⁴

Fitzpatrick skin type and/or minimal erythema dose testing also are essential in determining the appropriate initial NB-UVB dose required for treatment initiation (Table 2). Patient response to the initial NB-UVB trial will determine the optimal dosage for subsequent maintenance treatments.

Assessment and Optimization of Phototherapy

After an appropriate starting dosage has been established, patients should be evaluated at each subsequent visit for the degree of treatment response. Excessive erythema (lasting more than 48 hours) or adverse effects, such as itching, stinging, or burning, are indications that the patient should have their dose adjusted back to the last dose without such adverse responses. Because tolerance to treatment develops over time, patients who miss a scheduled dose of NB-UVB phototherapy require their dose to be temporarily lowered. Targeted dosage of UVB phototherapy at a frequency of 2 to 3 times weekly is preferred over treatment 1 to 2 times weekly; however, consideration should be given toward patient preference.²⁶ Dosing may be increased at a rate of 5% to 10% after each treatment, as tolerated, if there is no clearance of skin lesions with the original treatment dose. Patient skin type also is helpful in dictating the maximum phototherapy dose for each patient (Table 3).

Once a patient’s psoriatic lesions have cleared, the patient has the option to taper or indefinitely continue maintenance therapy. The established protocol for patients who choose to taper therapy is treatment twice weekly for 4 weeks, followed by once-weekly treatment for the second month. The maintenance dosage is held constant during the taper. For patients who prefer indefinite maintenance therapy, treatment is administered every 1 to 2 weeks, with a maintenance dosage that is approximately 25% lower than the original maintenance dosage.

Treatment Considerations

Efforts should be made to ensure that the long-term sequalae of phototherapy are minimized (Table 1). Development of cataracts is a recognized consequence of prolonged UVB exposure; therefore, eye protection is recommended during all UVB treatment sessions to reduce the risk for ocular toxicity. Although pregnancy is not a contraindication to phototherapy, except for PUVA, there is a dose-dependent degradation of folate with NB-UVB treatment, so folate supplementation (0.8 mg) is recommended during NB-UVB treatment to prevent development of neural tube defects in fetuses of patients who are pregnant or who may become pregnant.²⁷

Although phototherapy carries the theoretical risk for photocarcinogenesis, multiple studies have shown no increased risk for malignancy with either NB-UVB or BB-UVB phototherapy.^22,23 Regardless, patients who develop new-onset skin cancer while receiving any phototherapeutic treatment should discuss the potential risks and benefits of continued treatment with their physician. Providers should take extra caution prior to initiating treatment, especially in patients with a history of cutaneous malignancy. Because oral PUVA is a systemic therapy, it is associated with a greater risk for photocarcinogenesis than any other modality, particularly in fair-skinned individuals. Patients younger than 10 years; pregnant or nursing patients; and those with a history of lupus, xeroderma pigmentosum, or melanoma should not receive PUVA therapy because of their increased risk for photocarcinogenesis and TRAEs.

The decision to switch patients between different phototherapy modalities during treatment should be individualized to each patient based on factors such as disease severity, quality of life, and treatment burden. Other factors to consider include dosing frequency, treatment cost, and logistical factors, such as proximity to a treatment facility. Physicians should have a detailed discussion about the risks and benefits of continuing therapy for patients who develop new-onset skin cancer during phototherapy.

Final Thoughts

Phototherapy is an effective and safe treatment for patients with psoriasis who have localized and systemic disease. There are several treatment modalities that can be tailored to patient needs and preferences, such as home NB-UVB for patients who are unable or unwilling to undergo office-based treatments. Phototherapy has few absolute contraindications; however, relative contraindications to phototherapy exist. Patients with a history of skin cancer, photosensitivity disorders, and autoimmune diseases (eg, lupus) carry greater risks for adverse events, such as sun-related damage, cancer, and dysplasia. Because these conditions may preclude patients from pursuing phototherapy as a safe and effective approach to treating moderate to severe psoriasis, these patients should be considered for other therapies, such as biologic medications, which may carry a more favorable risk-benefit ratio given that individual’s background.

Psoriasis is a systemic immune-mediated disorder characterized by erythematous, scaly, well-demarcated plaques on the skin that affects approximately 3% of the world’s population.¹ Although topical therapies often are the first-line treatment of mild to moderate psoriasis, approximately 1 in 6 individuals has moderate to severe disease that requires systemic treatment such as biologics or phototherapy.² In patients with localized disease that is refractory to treatment or who have moderate to severe psoriasis requiring systemic treatment, phototherapy should be considered as a potential low-risk treatment option.

In July 2019, the American Academy of Dermatology (AAD) and National Psoriasis Foundation (NPF) released an updated set of guidelines for the use of phototherapy in treating adult patients with psoriasis.³ Since the prior guidelines were released in 2010, there have been numerous studies affirming the efficacy of phototherapy, with several large meta-analyses helping to refine clinical recommendations.^4,5 Each treatment was ranked using Strength of Recommendation Taxonomy, with a score of A, B, or C based on the strength of the evidence supporting the given modality. With the ever-increasing number of treatment options for patients with psoriasis, these guidelines inform dermatologists of the recommendations for the initiation, maintenance, and optimization of phototherapy in the treatment of psoriasis.

The AAD-NPF recommendations discuss the mechanism of action, efficacy, safety, and frequency of adverse events of 10 commonly used phototherapy/photochemotherapy modalities. They also address dosing regimens, the potential to combine phototherapy with other therapies, and the efficacy of treatment modalities for different types of psoriasis.³ The purpose of this discussion is to present these guidelines in a condensed form for prescribers of phototherapy and to review the most clinically significant considerations during each step of treatment. Of note, we only highlight the treatment of adult patients and do not discuss information relevant to pediatric patients with psoriasis.

Choosing a Phototherapy Modality

Phototherapy may be considered for patients with psoriasis that affects more than 3% body surface area or for localized disease refractory to conventional treatments. UV light is believed to provide relief from psoriasis via multiple mechanisms, such as through favorable alterations in cytokine profiles, initiation of apoptosis, and local immunosupression.⁶ There is no single first-line phototherapeutic modality recommended for all patients with psoriasis. Rather, the decision to implement a particular modality should be individualized to the patient, considering factors such as percentage of body surface area affected by disease, quality-of-life assessment, comorbidities, lifestyle, and cost of treatment.

Of the 10 phototherapy modalities reviewed in these guidelines, 4 were ranked by the AAD and NPF as having grade A evidence for efficacy in the treatment of moderate to severe plaque psoriasis. Treatments with a grade A level of recommendation included narrowband UVB (NB-UVB), broadband UVB (BB-UVB), targeted phototherapy (excimer laser and excimer lamp), and oral psoralen plus UVA (PUVA) therapy. Photodynamic therapy for psoriasis was given an A-level recommendation against its use, as it was found to be ineffective with an unfavorable side-effect profile. Treatments with a grade B level of recommendation—nonoral routes of PUVA therapy, pulsed dye laser/intense pulsed light for nail psoriasis only, Goeckerman therapy, and climatotherapy—have sufficient evidence available to support their treatment of moderate to severe psoriasis in some cases. Treatments with a grade C level of recommendation—Grenz ray therapy (also called borderline or ultrasoft therapy) and visible light therapy—have insufficient evidence to support their use in patients with moderate to severe psoriasis (Table 1).

Psoralen plus UVA, which may be used topically (ie, bathwater PUVA) or taken orally, refers to treatment with photosensitizing psoralens. Psoralens are agents that intercalate with DNA and enhance the efficacy of phototherapy.¹⁰ Topical PUVA, with a grade B level of recommendation, is an effective treatment option for patients with localized disease and has been shown to be particularly efficacious in the treatment of palmoplantar pustular psoriasis. Oral PUVA is an effective option for psoriasis with a grade A recommendation, while bathwater PUVA has a grade B level of recommendation for moderate to severe plaque psoriasis. Oral PUVA is associated with greater systemic side effects (both acute and subacute) compared with NB-UVB and also is associated with photocarcinogenesis, particularly squamous cell carcinoma in white patients.¹¹ Other side effects from PUVA include pigmented macules in sun-protected areas (known as PUVA lentigines), which may make evaluation of skin lesions challenging. Because of the increased risk for cancer with oral PUVA, NB-UVB is preferable as a first-line treatment vs PUVA, especially in patients with a history of skin cancer.^12,13

Goeckerman therapy, which involves the synergistic combination of UVB and crude coal tar, is an older treatment that has shown efficacy in the treatment of severe or recalcitrant psoriasis (grade B level of recommendation). One prior case-control study comparing the efficacy of Goeckerman therapy with newer treatments, such as biologic therapies, steroids, and oral immunosuppressants, found a similar reduction in symptoms among both treatment groups, with longer disease-free periods in patients who received Goeckerman therapy than those who received newer therapies (22.3 years vs 4.6 months).¹⁴ However, Goeckerman therapy is utilized less frequently than more modern therapies because of the time required for treatment and declining insurance reimbursements for it. Climatotherapy, another older established therapy, involves the temporary or permanent relocation of patients to an environment that is favorable for disease control (grade B level of recommendation). Locations such as the Dead Sea and Canary Islands have been studied and shown to provide both subjective and objective improvement in patients’ psoriasis disease course. Patients had notable improvement in both their psoriasis area and severity index score and quality of life after a 3- to 4-week relocation to these areas.^15,16 Access to climatotherapy and the transient nature of disease relief are apparent limitations of this treatment modality.

Grenz ray is a type of phototherapy that uses longer-wavelength ionizing radiation, which has low penetrance into surrounding tissues and a 95% absorption rate within the first 3 mm of the skin depth. This treatment has been used less frequently since the development of newer alternatives but should still be considered as a second line to UV therapy, especially in cases of recalcitrant disease and palmoplantar psoriasis, and when access to other treatment options is limited. Grenz ray has a grade C level of recommendation due to the paucity of evidence that supports its efficacy. Thus, it is not recommended as first-line therapy for the treatment of moderate to severe psoriasis. Visible light therapy is another treatment option that uses light in the visible wavelength spectrum but predominantly utilizes blue and red light. Psoriatic lesions are sensitive to light therapy because of the elevated levels of naturally occurring photosensitizing agents, called protoporphyrins, in these lesions.¹⁷ Several small studies have shown improvement in psoriatic lesions treated with visible light therapy, with blue light showing greater efficacy in lesional clearance than red light.^18,19

Pulsed dye laser is a phototherapy modality that has shown efficacy in the treatment of nail psoriasis (grade B level of recommendation). One study comparing the effects of tazarotene cream 0.1% with pulsed dye laser and tazarotene cream 0.1% alone showed that patients receiving combination therapy had a greater decrease in nail psoriasis severity index scores, higher scores on the patient’s global assessment of improvement, and higher rates of improvement on the physician global assessment score. A physician global assessment score of 75% improvement or more was seen in patients treated with combination therapy vs monotherapy (5.3% vs 31.6%).²⁰ Intense pulsed light, a type of visible light therapy, also has been used to treat nail psoriasis, with one study showing notable improvement in nail bed and matrix disease and a global improvement in nail psoriasis severity index score after 6 months of biweekly treatment.²¹ However, this treatment has a grade B level of recommendation given the limited number of studies supporting the efficacy of this modality.

Initiation of Phototherapy

Prior to initiating phototherapy, it is important to assess the patient for any personal or family history of skin cancer, as phototherapy carries an increased risk for cutaneous malignancy, especially in patients with a history of skin cancer.^22,23 All patients also should be evaluated for their Fitzpatrick skin type, and the minimal erythema dose should be defined for those initiating UVB treatment. These classifications can be useful for the initial determination of treatment dose and the prevention of treatment-related adverse events (TRAEs). A careful drug history also should be taken before the initiation of phototherapy to avoid photosensitizing reactions. Thiazide diuretics and tetracyclines are 2 such examples of medications commonly associated with photosensitizing reactions.²⁴

Fitzpatrick skin type and/or minimal erythema dose testing also are essential in determining the appropriate initial NB-UVB dose required for treatment initiation (Table 2). Patient response to the initial NB-UVB trial will determine the optimal dosage for subsequent maintenance treatments.

Assessment and Optimization of Phototherapy

After an appropriate starting dosage has been established, patients should be evaluated at each subsequent visit for the degree of treatment response. Excessive erythema (lasting more than 48 hours) or adverse effects, such as itching, stinging, or burning, are indications that the patient should have their dose adjusted back to the last dose without such adverse responses. Because tolerance to treatment develops over time, patients who miss a scheduled dose of NB-UVB phototherapy require their dose to be temporarily lowered. Targeted dosage of UVB phototherapy at a frequency of 2 to 3 times weekly is preferred over treatment 1 to 2 times weekly; however, consideration should be given toward patient preference.²⁶ Dosing may be increased at a rate of 5% to 10% after each treatment, as tolerated, if there is no clearance of skin lesions with the original treatment dose. Patient skin type also is helpful in dictating the maximum phototherapy dose for each patient (Table 3).

Once a patient’s psoriatic lesions have cleared, the patient has the option to taper or indefinitely continue maintenance therapy. The established protocol for patients who choose to taper therapy is treatment twice weekly for 4 weeks, followed by once-weekly treatment for the second month. The maintenance dosage is held constant during the taper. For patients who prefer indefinite maintenance therapy, treatment is administered every 1 to 2 weeks, with a maintenance dosage that is approximately 25% lower than the original maintenance dosage.

Treatment Considerations

Efforts should be made to ensure that the long-term sequalae of phototherapy are minimized (Table 1). Development of cataracts is a recognized consequence of prolonged UVB exposure; therefore, eye protection is recommended during all UVB treatment sessions to reduce the risk for ocular toxicity. Although pregnancy is not a contraindication to phototherapy, except for PUVA, there is a dose-dependent degradation of folate with NB-UVB treatment, so folate supplementation (0.8 mg) is recommended during NB-UVB treatment to prevent development of neural tube defects in fetuses of patients who are pregnant or who may become pregnant.²⁷

Although phototherapy carries the theoretical risk for photocarcinogenesis, multiple studies have shown no increased risk for malignancy with either NB-UVB or BB-UVB phototherapy.^22,23 Regardless, patients who develop new-onset skin cancer while receiving any phototherapeutic treatment should discuss the potential risks and benefits of continued treatment with their physician. Providers should take extra caution prior to initiating treatment, especially in patients with a history of cutaneous malignancy. Because oral PUVA is a systemic therapy, it is associated with a greater risk for photocarcinogenesis than any other modality, particularly in fair-skinned individuals. Patients younger than 10 years; pregnant or nursing patients; and those with a history of lupus, xeroderma pigmentosum, or melanoma should not receive PUVA therapy because of their increased risk for photocarcinogenesis and TRAEs.

The decision to switch patients between different phototherapy modalities during treatment should be individualized to each patient based on factors such as disease severity, quality of life, and treatment burden. Other factors to consider include dosing frequency, treatment cost, and logistical factors, such as proximity to a treatment facility. Physicians should have a detailed discussion about the risks and benefits of continuing therapy for patients who develop new-onset skin cancer during phototherapy.

Final Thoughts

Phototherapy is an effective and safe treatment for patients with psoriasis who have localized and systemic disease. There are several treatment modalities that can be tailored to patient needs and preferences, such as home NB-UVB for patients who are unable or unwilling to undergo office-based treatments. Phototherapy has few absolute contraindications; however, relative contraindications to phototherapy exist. Patients with a history of skin cancer, photosensitivity disorders, and autoimmune diseases (eg, lupus) carry greater risks for adverse events, such as sun-related damage, cancer, and dysplasia. Because these conditions may preclude patients from pursuing phototherapy as a safe and effective approach to treating moderate to severe psoriasis, these patients should be considered for other therapies, such as biologic medications, which may carry a more favorable risk-benefit ratio given that individual’s background.

Psoriasis is a systemic immune-mediated disorder characterized by erythematous, scaly, well-demarcated plaques on the skin that affects approximately 3% of the world’s population.¹ Although topical therapies often are the first-line treatment of mild to moderate psoriasis, approximately 1 in 6 individuals has moderate to severe disease that requires systemic treatment such as biologics or phototherapy.² In patients with localized disease that is refractory to treatment or who have moderate to severe psoriasis requiring systemic treatment, phototherapy should be considered as a potential low-risk treatment option.

In July 2019, the American Academy of Dermatology (AAD) and National Psoriasis Foundation (NPF) released an updated set of guidelines for the use of phototherapy in treating adult patients with psoriasis.³ Since the prior guidelines were released in 2010, there have been numerous studies affirming the efficacy of phototherapy, with several large meta-analyses helping to refine clinical recommendations.^4,5 Each treatment was ranked using Strength of Recommendation Taxonomy, with a score of A, B, or C based on the strength of the evidence supporting the given modality. With the ever-increasing number of treatment options for patients with psoriasis, these guidelines inform dermatologists of the recommendations for the initiation, maintenance, and optimization of phototherapy in the treatment of psoriasis.

The AAD-NPF recommendations discuss the mechanism of action, efficacy, safety, and frequency of adverse events of 10 commonly used phototherapy/photochemotherapy modalities. They also address dosing regimens, the potential to combine phototherapy with other therapies, and the efficacy of treatment modalities for different types of psoriasis.³ The purpose of this discussion is to present these guidelines in a condensed form for prescribers of phototherapy and to review the most clinically significant considerations during each step of treatment. Of note, we only highlight the treatment of adult patients and do not discuss information relevant to pediatric patients with psoriasis.

Choosing a Phototherapy Modality

Phototherapy may be considered for patients with psoriasis that affects more than 3% body surface area or for localized disease refractory to conventional treatments. UV light is believed to provide relief from psoriasis via multiple mechanisms, such as through favorable alterations in cytokine profiles, initiation of apoptosis, and local immunosupression.⁶ There is no single first-line phototherapeutic modality recommended for all patients with psoriasis. Rather, the decision to implement a particular modality should be individualized to the patient, considering factors such as percentage of body surface area affected by disease, quality-of-life assessment, comorbidities, lifestyle, and cost of treatment.

Of the 10 phototherapy modalities reviewed in these guidelines, 4 were ranked by the AAD and NPF as having grade A evidence for efficacy in the treatment of moderate to severe plaque psoriasis. Treatments with a grade A level of recommendation included narrowband UVB (NB-UVB), broadband UVB (BB-UVB), targeted phototherapy (excimer laser and excimer lamp), and oral psoralen plus UVA (PUVA) therapy. Photodynamic therapy for psoriasis was given an A-level recommendation against its use, as it was found to be ineffective with an unfavorable side-effect profile. Treatments with a grade B level of recommendation—nonoral routes of PUVA therapy, pulsed dye laser/intense pulsed light for nail psoriasis only, Goeckerman therapy, and climatotherapy—have sufficient evidence available to support their treatment of moderate to severe psoriasis in some cases. Treatments with a grade C level of recommendation—Grenz ray therapy (also called borderline or ultrasoft therapy) and visible light therapy—have insufficient evidence to support their use in patients with moderate to severe psoriasis (Table 1).

Psoralen plus UVA, which may be used topically (ie, bathwater PUVA) or taken orally, refers to treatment with photosensitizing psoralens. Psoralens are agents that intercalate with DNA and enhance the efficacy of phototherapy.¹⁰ Topical PUVA, with a grade B level of recommendation, is an effective treatment option for patients with localized disease and has been shown to be particularly efficacious in the treatment of palmoplantar pustular psoriasis. Oral PUVA is an effective option for psoriasis with a grade A recommendation, while bathwater PUVA has a grade B level of recommendation for moderate to severe plaque psoriasis. Oral PUVA is associated with greater systemic side effects (both acute and subacute) compared with NB-UVB and also is associated with photocarcinogenesis, particularly squamous cell carcinoma in white patients.¹¹ Other side effects from PUVA include pigmented macules in sun-protected areas (known as PUVA lentigines), which may make evaluation of skin lesions challenging. Because of the increased risk for cancer with oral PUVA, NB-UVB is preferable as a first-line treatment vs PUVA, especially in patients with a history of skin cancer.^12,13

Goeckerman therapy, which involves the synergistic combination of UVB and crude coal tar, is an older treatment that has shown efficacy in the treatment of severe or recalcitrant psoriasis (grade B level of recommendation). One prior case-control study comparing the efficacy of Goeckerman therapy with newer treatments, such as biologic therapies, steroids, and oral immunosuppressants, found a similar reduction in symptoms among both treatment groups, with longer disease-free periods in patients who received Goeckerman therapy than those who received newer therapies (22.3 years vs 4.6 months).¹⁴ However, Goeckerman therapy is utilized less frequently than more modern therapies because of the time required for treatment and declining insurance reimbursements for it. Climatotherapy, another older established therapy, involves the temporary or permanent relocation of patients to an environment that is favorable for disease control (grade B level of recommendation). Locations such as the Dead Sea and Canary Islands have been studied and shown to provide both subjective and objective improvement in patients’ psoriasis disease course. Patients had notable improvement in both their psoriasis area and severity index score and quality of life after a 3- to 4-week relocation to these areas.^15,16 Access to climatotherapy and the transient nature of disease relief are apparent limitations of this treatment modality.

Grenz ray is a type of phototherapy that uses longer-wavelength ionizing radiation, which has low penetrance into surrounding tissues and a 95% absorption rate within the first 3 mm of the skin depth. This treatment has been used less frequently since the development of newer alternatives but should still be considered as a second line to UV therapy, especially in cases of recalcitrant disease and palmoplantar psoriasis, and when access to other treatment options is limited. Grenz ray has a grade C level of recommendation due to the paucity of evidence that supports its efficacy. Thus, it is not recommended as first-line therapy for the treatment of moderate to severe psoriasis. Visible light therapy is another treatment option that uses light in the visible wavelength spectrum but predominantly utilizes blue and red light. Psoriatic lesions are sensitive to light therapy because of the elevated levels of naturally occurring photosensitizing agents, called protoporphyrins, in these lesions.¹⁷ Several small studies have shown improvement in psoriatic lesions treated with visible light therapy, with blue light showing greater efficacy in lesional clearance than red light.^18,19

Pulsed dye laser is a phototherapy modality that has shown efficacy in the treatment of nail psoriasis (grade B level of recommendation). One study comparing the effects of tazarotene cream 0.1% with pulsed dye laser and tazarotene cream 0.1% alone showed that patients receiving combination therapy had a greater decrease in nail psoriasis severity index scores, higher scores on the patient’s global assessment of improvement, and higher rates of improvement on the physician global assessment score. A physician global assessment score of 75% improvement or more was seen in patients treated with combination therapy vs monotherapy (5.3% vs 31.6%).²⁰ Intense pulsed light, a type of visible light therapy, also has been used to treat nail psoriasis, with one study showing notable improvement in nail bed and matrix disease and a global improvement in nail psoriasis severity index score after 6 months of biweekly treatment.²¹ However, this treatment has a grade B level of recommendation given the limited number of studies supporting the efficacy of this modality.

Initiation of Phototherapy

Prior to initiating phototherapy, it is important to assess the patient for any personal or family history of skin cancer, as phototherapy carries an increased risk for cutaneous malignancy, especially in patients with a history of skin cancer.^22,23 All patients also should be evaluated for their Fitzpatrick skin type, and the minimal erythema dose should be defined for those initiating UVB treatment. These classifications can be useful for the initial determination of treatment dose and the prevention of treatment-related adverse events (TRAEs). A careful drug history also should be taken before the initiation of phototherapy to avoid photosensitizing reactions. Thiazide diuretics and tetracyclines are 2 such examples of medications commonly associated with photosensitizing reactions.²⁴

Fitzpatrick skin type and/or minimal erythema dose testing also are essential in determining the appropriate initial NB-UVB dose required for treatment initiation (Table 2). Patient response to the initial NB-UVB trial will determine the optimal dosage for subsequent maintenance treatments.

Assessment and Optimization of Phototherapy

After an appropriate starting dosage has been established, patients should be evaluated at each subsequent visit for the degree of treatment response. Excessive erythema (lasting more than 48 hours) or adverse effects, such as itching, stinging, or burning, are indications that the patient should have their dose adjusted back to the last dose without such adverse responses. Because tolerance to treatment develops over time, patients who miss a scheduled dose of NB-UVB phototherapy require their dose to be temporarily lowered. Targeted dosage of UVB phototherapy at a frequency of 2 to 3 times weekly is preferred over treatment 1 to 2 times weekly; however, consideration should be given toward patient preference.²⁶ Dosing may be increased at a rate of 5% to 10% after each treatment, as tolerated, if there is no clearance of skin lesions with the original treatment dose. Patient skin type also is helpful in dictating the maximum phototherapy dose for each patient (Table 3).

Once a patient’s psoriatic lesions have cleared, the patient has the option to taper or indefinitely continue maintenance therapy. The established protocol for patients who choose to taper therapy is treatment twice weekly for 4 weeks, followed by once-weekly treatment for the second month. The maintenance dosage is held constant during the taper. For patients who prefer indefinite maintenance therapy, treatment is administered every 1 to 2 weeks, with a maintenance dosage that is approximately 25% lower than the original maintenance dosage.

Treatment Considerations

Efforts should be made to ensure that the long-term sequalae of phototherapy are minimized (Table 1). Development of cataracts is a recognized consequence of prolonged UVB exposure; therefore, eye protection is recommended during all UVB treatment sessions to reduce the risk for ocular toxicity. Although pregnancy is not a contraindication to phototherapy, except for PUVA, there is a dose-dependent degradation of folate with NB-UVB treatment, so folate supplementation (0.8 mg) is recommended during NB-UVB treatment to prevent development of neural tube defects in fetuses of patients who are pregnant or who may become pregnant.²⁷

Although phototherapy carries the theoretical risk for photocarcinogenesis, multiple studies have shown no increased risk for malignancy with either NB-UVB or BB-UVB phototherapy.^22,23 Regardless, patients who develop new-onset skin cancer while receiving any phototherapeutic treatment should discuss the potential risks and benefits of continued treatment with their physician. Providers should take extra caution prior to initiating treatment, especially in patients with a history of cutaneous malignancy. Because oral PUVA is a systemic therapy, it is associated with a greater risk for photocarcinogenesis than any other modality, particularly in fair-skinned individuals. Patients younger than 10 years; pregnant or nursing patients; and those with a history of lupus, xeroderma pigmentosum, or melanoma should not receive PUVA therapy because of their increased risk for photocarcinogenesis and TRAEs.

The decision to switch patients between different phototherapy modalities during treatment should be individualized to each patient based on factors such as disease severity, quality of life, and treatment burden. Other factors to consider include dosing frequency, treatment cost, and logistical factors, such as proximity to a treatment facility. Physicians should have a detailed discussion about the risks and benefits of continuing therapy for patients who develop new-onset skin cancer during phototherapy.

Final Thoughts

Phototherapy is an effective and safe treatment for patients with psoriasis who have localized and systemic disease. There are several treatment modalities that can be tailored to patient needs and preferences, such as home NB-UVB for patients who are unable or unwilling to undergo office-based treatments. Phototherapy has few absolute contraindications; however, relative contraindications to phototherapy exist. Patients with a history of skin cancer, photosensitivity disorders, and autoimmune diseases (eg, lupus) carry greater risks for adverse events, such as sun-related damage, cancer, and dysplasia. Because these conditions may preclude patients from pursuing phototherapy as a safe and effective approach to treating moderate to severe psoriasis, these patients should be considered for other therapies, such as biologic medications, which may carry a more favorable risk-benefit ratio given that individual’s background.

From the Department of Medicine, Division of Cardiology, Northwestern University Feinberg School of Medicine, Chicago, IL.

Abstract

Objective: To review selected process-of-care interventions that can be applied both during the hospitalization and during the transitional care period to help address the persistent challenge of heart failure readmissions.

Methods: Review of the literature.

Results: Process-of-care interventions that can be implemented to reduce readmissions of heart failure patients include: accurately identifying heart failure patients; providing disease education; titrating guideline-directed medical therapy; ensuring discharge readiness; arranging close discharge follow-up; identifying and addressing social barriers; following up by telephone; using home health; and addressing comorbidities. Importantly, the heart failure hospitalization is an opportunity to set up outpatient success, and setting up feedback loops can aid in post-discharge monitoring.

Conclusion: We encourage teams to consider local capabilities when selecting processes to improve; begin by improving something small to build capacity and team morale, and continually iterate and reexamine processes, as health care systems are continually evolving.

Keywords: heart failure; process improvement; quality improvement; readmission; rehospitalization; transitional care.

The growing population of patients affected by heart failure continues to challenge health systems. The increasing prevalence is paralleled by the rising costs of managing heart failure, which are projected to grow from $30.7 billion in 2012 to $69.8 billion in 2030.¹ A significant portion of these costs relate to readmission after an index heart failure hospitalization. The statistics are staggering: for patients hospitalized with heart failure, approximately 15% to 20% are readmitted within 30 days.^2,3 Though recent temporal trends suggest a modest reduction in readmission rates, there is a concerning correlation with increasing mortality,³ and a recognition that readmission rate decreases may relate to subtle changes in coding-based risk adjustment.⁴ Despite these concerns, efforts to reduce readmissions after heart failure hospitalization command significant attention.

Process improvement methodologies may be helpful in reducing hospital readmissions. Various approaches have been employed, and results have been mixed. An analysis of 70 participating hospitals in the American Heart Association’s Get With the Guidelines initiative found that, while overall readmission rates declined by 1.0% over 3 years, only 1 hospital achieved a 20% reduction in readmission rates.⁵

It is notably difficult to reduce readmissions after heart failure hospitalization. One challenge is that patients with heart failure often have multiple comorbidities, and approximately 50% to 60% of 30-day readmissions after heart failure hospitalization arise from noncardiac causes.¹ Another challenge is that a significant fraction of readmissions in general—perhaps 75%—may not be avoidable.⁶

Recent excellent systematic reviews and meta-analyses provide comprehensive overviews of process improvement strategies that can be used to reduce readmissions after heart failure hospitalizations.^7-9 Yet despite this extensive knowledge, few reports discuss the process of actually implementing these changes: the process of process improvement. Here, we seek to not only highlight some of the most promising potential interventions to reduce heart failure readmissions, but also to discuss a process improvement framework to help engender success, using our experience as a case study. We schematize process improvement efforts as having several distinct phases (Figure 1): processes delivered during the hospitalization and prior to discharge; feedback loops set up to maintain clinical stability at home; and the postdischarge clinic visit as an opportunity to further stabilize the patient and advance the plan of care. The discussion of these interventions follows this organization.

During Hospitalization

The heart failure hospitalization can be used as an opportunity to set up outpatient success, with several goals to target during the index admission. One goal is identifying the root causes of the heart failure syndrome and correcting those root causes, if possible. For example, patients in whom the heart failure syndrome is secondary to valvular heart disease may benefit from transcatheter aortic valve replacement.¹⁰ Another clinical goal is decongesting the patient, which is associated with lower readmission rates.^11,12 These goals focus on the medical aspects of heart failure care. However, beyond these medical aspects, a patient must be equipped to successfully manage the disease at home.

To support medical and nonmedical interventions for hospitalized heart failure patients, a critical first step is identifying patients with heart failure. This accomplishes at least 2 objectives. First, early identification allows early initiation of interventions, such as heart failure education and social work evaluation. Early initiation of these interventions allows sufficient time during the hospitalization to make meaningful progress on these fronts. Second, early identification allows an opportunity for the delivery of cardiology specialty care, which may help with identifying and correcting root causes of the heart failure syndrome. Such access to cardiology has been shown to improve inpatient mortality and readmission rates.¹³

In smaller hospitals, identification of patients with heart failure can be as simple as reviewing overnight admissions. More advanced strategies, such as screeners based on brain natriuretic peptide (BNP) levels and administration of intravenous diuretics, can be employed.^14,15 In the near future, deep learning-based natural language processing will be applied to mine full-text data in the electronic health record to identify heart failure hospitalizations.¹⁶

In the hospital, patients can also receive education about heart failure disease management. This education is a cornerstone of reducing heart failure readmissions. A recent systematic review of nurse education interventions demonstrated reductions in readmissions, hospitalizations, and costs.¹⁷ However, the efficacy of heart failure education hinges on many other variables. For patients to adhere to water restriction and daily weights, for example, there must also be patient understanding, compliance, and accessibility to providers to recommend how to strike the fluid balance. Education is therefore necessary, but not sufficient, for setting up outpatient success.

The hospitalization also represents an important time to start or uptitrate guideline-directed medical therapy (GDMT) for heart failure. Doing so takes advantage of an important opportunity to reduce the risk of readmission and even reverse the disease process.¹⁸ Uptitration of GDMT in patients with heart failure with reduced ejection fraction is associated with a decreased risk of mortality, while discontinuation is associated with an increased risk of mortality.¹⁹ However, recent registry data indicate that intensity of GDMT is just as likely to be decreased as increased during the hospitalization.²⁰ Nevertheless, predischarge initiation of medications may be associated with higher attained doses in follow-up.²¹

Preparing for Discharge

Preparing a patient for discharge after a heart failure hospitalization involves stabilizing the medical condition as well as ensuring that the patient and caregivers have the medication, equipment, and self-care resources at home necessary to manage the condition. Several frameworks have been put forth to help care teams analyze a patient’s readiness for discharge. One is the B-PREPARED score,²² a validated instrument to discriminate among patients with regard to their readiness to discharge from the hospital. This instrument highlights the importance of several key factors that should be addressed during the discharge process, including counseling and written instructions about medications and their side effects; information about equipment needs and community resources; and information on activity levels and restrictions. Nurse education and discharge coordination can improve patients’ perception of discharge readiness,²³ although whether this discharge readiness translates into improved readmission rates appears to depend on the specific follow-up intervention design.⁹

Prior to discharge, it is important to arrange postdischarge follow-up appointments, as emphasized by the American College of Cardiology/American Heart Association (ACC/AHA) guidelines.²⁴ The use of nurse navigators can help with planning follow-up appointments. For example, the ACC Patient Navigator Program was applied in a single-center study of 120 patients randomized to the program versus usual care.²⁵ This study found a significant increase in patient education and follow-up appointments compared to usual care, and a numerical decrease in hospital readmissions, although the finding was not statistically significant.²⁵

A third critical component of preparing for discharge is identifying and addressing social barriers to care. In a study of patients stratified by household income, patients in the lowest income quartile had a higher readmission rate than patients in the highest income quartile.²⁶ Poverty also correlates with heart failure mortality.²⁷ Social factors play an important role in many aspects of patients’ ability to manage their health, including self-care, medication adherence, and ability to follow-up. Identifying these social factors prior to discharge is the first step to addressing them. While few studies specifically address the role of social workers in the management of heart failure care, the general medical literature suggests that social workers embedded in transitional care teams can augment readmission reduction efforts.²⁸

After Discharge

Patients recently discharged from the hospital who have not yet attended their postdischarge appointment are in an incredibly vulnerable phase of care. Patients who are discharged from the hospital may not yet be connected with outpatient care. During this initial transitional care period, feedback loops involving patient communication back to the clinic, and clinic communication back to the patient, are critical to helping patients remain stable. For example, consider monitoring weights daily after hospital discharge. A patient at home can report increasing weights to a provider, who can then recommend an increased dose of diuretic. The patient can complete the feedback loop by taking the extra medication and monitoring the return of weight back to normal.

While daily weight monitoring is a simple process improvement that relies on the principle of establishing feedback loops, many other strategies exist. One commonly employed tool is the postdischarge telephone follow-up call, which is often coupled with other interventions in a comprehensive care bundle.⁸ During the telephone call, several process-of-care defects can be corrected, including missing medications or missing information on appointment times.

Beyond the telephone, newer technologies show promise for helping develop feedback loops for patients at home. One such technology is telemonitoring, whereby physiologic information such as weight, heart rate, and blood pressure is collected and sent back to a monitoring center. While the principle holds promise, several studies have not demonstrated significantly different outcomes as compared to usual care.^13,29 Another promising technology is the CardioMEMS device (Abbott, Inc., Atlanta, GA), which can remotely transmit the pulmonary artery pressure, a physiologic signal which correlates with volume overload. There is now strong evidence supporting the efficacy of pulmonary artery pressure–guided heart failure management.^30,31

Finally, home visits can be an efficient way to communicate symptoms, enable clinical assessment, and provide recommendations. One program that implemented home visits, 24-hour nurses available by call, and telephone follow-up showed a statistically significant reduction in readmissions.³² Furthermore, a meta-analysis of randomized controlled trials comparing home health to usual care showed decreased readmissions and mortality.³³ The efficacy may be in strengthening the feedback loop—home care improves compliance with weight monitoring, fluid restriction, and medications.³⁴ These studies provide a strong rationale for the benefits of home health in stabilizing heart failure patients postdischarge. Indeed, nurse home visits were 1 of the 2 process interventions in a Cochrane review of randomized controlled trials that were shown to statistically significantly decrease readmissions and mortality.⁹ These data underscore the importance of feedback loops for helping ensure patients are clinically stable.

Postdischarge Follow-Up Clinic Visit

The first clinic appointment postdischarge is an important check-in to help advance patient care. Several key tasks can be achieved during the postdischarge visit. First, the patient can be clinically stabilized by adjusting diuretic therapy. If the patient is clinically stable, GDMT can be uptitrated. Second, education around symptoms, medications, diet, and exercise can be reinforced. Finally, clinicians can help connect patients to other members of the multidisciplinary care team, including specialist care, home health, or cardiac rehabilitation.

Achieving 7-day follow-up visits after discharge has been a point of emphasis in national guidelines.²⁴ The ACC promotes a “See You in 7” challenge, advising that all patients discharged with a diagnosis of heart failure have a follow-up appointment within 7 days. Yet based on the latest available data, arrival rates to the postdischarge clinic are dismal, hovering around 30%.³⁵ In a multicenter observational study of hospitals participating in the “See You in 7” collaborative, hospitals were able to increase their 7-day follow-up appointment rates by 2% to 3%, and also noted an absolute decrease in readmission rates by 1% to 2%.³⁶ We have demonstrated, using a mathematical approach called queuing theory, that discharge appointment wait times and clinic access can be significantly improved by providing a modest capacity buffer to clinic availability.³⁷ Those interested in applying this model to their own clinical practice may do so with a free online calculator at http://hfresearch.org.

Another important aspect of postdischarge follow-up is appropriate management of the comorbidity burden, which, as noted, is often significant in patients hospitalized with heart failure.³⁸ For instance, in recent cohorts of hospitalized heart failure patients, the incidence of hypertension was 78%, coronary artery disease was more than 50%, atrial fibrillation was more than 40%, and diabetes was nearly 40%.³⁹ Given this burden of comorbidity, it is not surprising that only 35% of readmissions after an index heart failure hospitalization are for recurrent heart failure.⁴⁰ Coordinating care among primary care physicians and relevant subspecialists is thus essential. Phone calls and secure electronic messages are very helpful in achieving this. There is increasing interest in more nimble care models, such as the patient-centered specialty practice⁴¹ or the dyspnea clinic, to help bring coordinated resources to the patient.⁴²

Process of Process Improvement: Our Experiences

The previous sections outline a series of potential process improvements clinical teams and health systems can implement to impact heart failure readmissions. A plan on paper, however, does not equal a plan in actuality. How does one go about implementing these changes? We offer our local experience starting a heart failure transitional care program as a case study, then draw lessons learned as a set of practical tips for local teams to employ. What we hope to highlight is that there is a large difference between a completed process for transitional care of heart failure patients, and the process of developing that process itself. The former is the hardware, the latter is the software. The latter does not typically get highlighted, but it is absolutely critical to unlocking the capabilities of a team and the institution.

In 2015, Northwestern Memorial Hospital adopted a novel payment arrangement from the Center for Medicare and Medicaid Services for Medicare patients being discharged from the hospital with heart failure. Known as Bundled Payments for Care Improvement,⁴³ this bundled payment model incentivized Northwestern Memorial Hospital charge, principally by reducing hospital readmissions and by collaborating with skilled nursing facilities to control length of stay.

We approached this problem by drawing on the available literature,^44,45 and by first creating a schematic of our high-level approach, which comprised 3 major elements (Figure 2): identification of hospitalized heart failure patients, delivery of a care bundle to hospitalized heart failure patients in hospital, and coordinating postdischarge care, centered on a telephone call and a postdischarge visit.

We then proceeded by building out, in stepwise fashion, each component of our value chain, using Agile techniques as a guiding principle.⁴⁶ Agile, a productivity and process improvement mindset with roots in software development, emphasizes tackling 1 problem at a time, building out new features sequentially and completely, recognizing that the end user does not derive value from a program until new functionality is available for use. Rather than wholesale monolithic change, Agile emphasizes rapid iteration, prototyping, and discarding innovations not found to be helpful. The notion is to stand up new, incremental features rapidly, with each incremental improvement delivering value and helping to accelerate overall change.

Our experience building a robust way to identify heart failure cases is a good example of Agile process improvement in practice. At our hospital, identification of patients with heart failure was a challenge because more than half of heart failure patients are admitted to noncardiology floors. We developed a simple electronic health record query to detect heart failure patients, relying on parameters such as administration of intravenous diuretic or levels of BNP exceeding 100 ng/dL. We deployed this query, finding very high sensitivity for detection of heart failure patients.¹⁴ Patients found to have heart failure were then populated into a list in the electronic health record, which made patients’ heart failure status visible to all members of the health care team. Using this list, we were able to automate several processes necessary for heart failure care. For example, the list made it possible for cardiologists to know if there was a patient who perhaps needed cardiology consultation. Nurse navigators could know which patients needed heart failure education without having to be actively consulted by the admitting team. The same nurse navigators could then know upon discharge which patients needed a follow-up telephone call at 48 hours.

This list of heart failure patients was the end product, which was built through prototyping and iteration. For example, with our initial BNP cutoff of 300 ng/dL, we recognized we were missing several cases, and lowered the cutoff for the screener to 100 ng/dL. When we were satisfied this process was working well, we moved on to the next problem to tackle, avoiding trying to work on too many things at once. By doing so, we were able to focus our process improvement resources on 1 problem at a time, building up a suite of interventions. For our hospital, we settled on a bundle of interventions, captured by the mnemonic HEART:

Heart doctor sees patient in the hospital

Education about heart failure in the hospital

After-visit summary with 7-day appointment printed

Reach out to the patient by telephone within 72 hours

Treat the patient in clinic by the 7-day visit

Conclusion

We would like to emphasize that the elements of our heart failure readmissions interventions were not all put in place at once. This was an iterative process that proceeded in a stepwise fashion, with each step improving the care of our patients. We learned a number of lessons from our experience. First, we would advise that teams not try to do everything. One program simply cannot implement all possible readmission reduction interventions, and certainly not all at once. Trade-offs should be made, and interventions more likely to succeed in the local environment should be prioritized. In addition, interventions that do not fit and do not create synergy with the local practice environment should not be pursued.

Second, we would advise teams to start small, tackling a known problem in heart failure transitions of care first. This initial intuition is often right. An example might be improving 7-day appointments upon discharge. Starting with a problem that can be tackled builds process improvement muscle and improves team morale. Third, we would advise teams to consistently iterate on designs, tweaking and improving performance. Complex organizations always evolve; processes that work 1 year may fail the next because another element of the organization may have changed.

Finally, the framework presented in Figure 1 may be helpful in guiding how to structure interventions. Considering interventions to be delivered in the hospital, interventions to be delivered in the clinic, and how to set up feedback loops to support patients as outpatients help develop a comprehensive heart failure readmissions reduction program.

Corresponding author: R. Kannan Mutharasan, MD, Northwestern University Feinberg School of Medicine, 676 North Saint Clair St., Arkes Pavilion, Suite 7-038, Chicago, IL 60611;[email protected].

Financial disclosures: None.

From the Department of Medicine, Division of Cardiology, Northwestern University Feinberg School of Medicine, Chicago, IL.

Abstract

Objective: To review selected process-of-care interventions that can be applied both during the hospitalization and during the transitional care period to help address the persistent challenge of heart failure readmissions.

Methods: Review of the literature.

Results: Process-of-care interventions that can be implemented to reduce readmissions of heart failure patients include: accurately identifying heart failure patients; providing disease education; titrating guideline-directed medical therapy; ensuring discharge readiness; arranging close discharge follow-up; identifying and addressing social barriers; following up by telephone; using home health; and addressing comorbidities. Importantly, the heart failure hospitalization is an opportunity to set up outpatient success, and setting up feedback loops can aid in post-discharge monitoring.

Conclusion: We encourage teams to consider local capabilities when selecting processes to improve; begin by improving something small to build capacity and team morale, and continually iterate and reexamine processes, as health care systems are continually evolving.

Keywords: heart failure; process improvement; quality improvement; readmission; rehospitalization; transitional care.

The growing population of patients affected by heart failure continues to challenge health systems. The increasing prevalence is paralleled by the rising costs of managing heart failure, which are projected to grow from $30.7 billion in 2012 to $69.8 billion in 2030.¹ A significant portion of these costs relate to readmission after an index heart failure hospitalization. The statistics are staggering: for patients hospitalized with heart failure, approximately 15% to 20% are readmitted within 30 days.^2,3 Though recent temporal trends suggest a modest reduction in readmission rates, there is a concerning correlation with increasing mortality,³ and a recognition that readmission rate decreases may relate to subtle changes in coding-based risk adjustment.⁴ Despite these concerns, efforts to reduce readmissions after heart failure hospitalization command significant attention.

Process improvement methodologies may be helpful in reducing hospital readmissions. Various approaches have been employed, and results have been mixed. An analysis of 70 participating hospitals in the American Heart Association’s Get With the Guidelines initiative found that, while overall readmission rates declined by 1.0% over 3 years, only 1 hospital achieved a 20% reduction in readmission rates.⁵

It is notably difficult to reduce readmissions after heart failure hospitalization. One challenge is that patients with heart failure often have multiple comorbidities, and approximately 50% to 60% of 30-day readmissions after heart failure hospitalization arise from noncardiac causes.¹ Another challenge is that a significant fraction of readmissions in general—perhaps 75%—may not be avoidable.⁶

Recent excellent systematic reviews and meta-analyses provide comprehensive overviews of process improvement strategies that can be used to reduce readmissions after heart failure hospitalizations.^7-9 Yet despite this extensive knowledge, few reports discuss the process of actually implementing these changes: the process of process improvement. Here, we seek to not only highlight some of the most promising potential interventions to reduce heart failure readmissions, but also to discuss a process improvement framework to help engender success, using our experience as a case study. We schematize process improvement efforts as having several distinct phases (Figure 1): processes delivered during the hospitalization and prior to discharge; feedback loops set up to maintain clinical stability at home; and the postdischarge clinic visit as an opportunity to further stabilize the patient and advance the plan of care. The discussion of these interventions follows this organization.

During Hospitalization

The heart failure hospitalization can be used as an opportunity to set up outpatient success, with several goals to target during the index admission. One goal is identifying the root causes of the heart failure syndrome and correcting those root causes, if possible. For example, patients in whom the heart failure syndrome is secondary to valvular heart disease may benefit from transcatheter aortic valve replacement.¹⁰ Another clinical goal is decongesting the patient, which is associated with lower readmission rates.^11,12 These goals focus on the medical aspects of heart failure care. However, beyond these medical aspects, a patient must be equipped to successfully manage the disease at home.

To support medical and nonmedical interventions for hospitalized heart failure patients, a critical first step is identifying patients with heart failure. This accomplishes at least 2 objectives. First, early identification allows early initiation of interventions, such as heart failure education and social work evaluation. Early initiation of these interventions allows sufficient time during the hospitalization to make meaningful progress on these fronts. Second, early identification allows an opportunity for the delivery of cardiology specialty care, which may help with identifying and correcting root causes of the heart failure syndrome. Such access to cardiology has been shown to improve inpatient mortality and readmission rates.¹³

In smaller hospitals, identification of patients with heart failure can be as simple as reviewing overnight admissions. More advanced strategies, such as screeners based on brain natriuretic peptide (BNP) levels and administration of intravenous diuretics, can be employed.^14,15 In the near future, deep learning-based natural language processing will be applied to mine full-text data in the electronic health record to identify heart failure hospitalizations.¹⁶

In the hospital, patients can also receive education about heart failure disease management. This education is a cornerstone of reducing heart failure readmissions. A recent systematic review of nurse education interventions demonstrated reductions in readmissions, hospitalizations, and costs.¹⁷ However, the efficacy of heart failure education hinges on many other variables. For patients to adhere to water restriction and daily weights, for example, there must also be patient understanding, compliance, and accessibility to providers to recommend how to strike the fluid balance. Education is therefore necessary, but not sufficient, for setting up outpatient success.

The hospitalization also represents an important time to start or uptitrate guideline-directed medical therapy (GDMT) for heart failure. Doing so takes advantage of an important opportunity to reduce the risk of readmission and even reverse the disease process.¹⁸ Uptitration of GDMT in patients with heart failure with reduced ejection fraction is associated with a decreased risk of mortality, while discontinuation is associated with an increased risk of mortality.¹⁹ However, recent registry data indicate that intensity of GDMT is just as likely to be decreased as increased during the hospitalization.²⁰ Nevertheless, predischarge initiation of medications may be associated with higher attained doses in follow-up.²¹

Preparing for Discharge

Preparing a patient for discharge after a heart failure hospitalization involves stabilizing the medical condition as well as ensuring that the patient and caregivers have the medication, equipment, and self-care resources at home necessary to manage the condition. Several frameworks have been put forth to help care teams analyze a patient’s readiness for discharge. One is the B-PREPARED score,²² a validated instrument to discriminate among patients with regard to their readiness to discharge from the hospital. This instrument highlights the importance of several key factors that should be addressed during the discharge process, including counseling and written instructions about medications and their side effects; information about equipment needs and community resources; and information on activity levels and restrictions. Nurse education and discharge coordination can improve patients’ perception of discharge readiness,²³ although whether this discharge readiness translates into improved readmission rates appears to depend on the specific follow-up intervention design.⁹

Prior to discharge, it is important to arrange postdischarge follow-up appointments, as emphasized by the American College of Cardiology/American Heart Association (ACC/AHA) guidelines.²⁴ The use of nurse navigators can help with planning follow-up appointments. For example, the ACC Patient Navigator Program was applied in a single-center study of 120 patients randomized to the program versus usual care.²⁵ This study found a significant increase in patient education and follow-up appointments compared to usual care, and a numerical decrease in hospital readmissions, although the finding was not statistically significant.²⁵

A third critical component of preparing for discharge is identifying and addressing social barriers to care. In a study of patients stratified by household income, patients in the lowest income quartile had a higher readmission rate than patients in the highest income quartile.²⁶ Poverty also correlates with heart failure mortality.²⁷ Social factors play an important role in many aspects of patients’ ability to manage their health, including self-care, medication adherence, and ability to follow-up. Identifying these social factors prior to discharge is the first step to addressing them. While few studies specifically address the role of social workers in the management of heart failure care, the general medical literature suggests that social workers embedded in transitional care teams can augment readmission reduction efforts.²⁸

After Discharge

Patients recently discharged from the hospital who have not yet attended their postdischarge appointment are in an incredibly vulnerable phase of care. Patients who are discharged from the hospital may not yet be connected with outpatient care. During this initial transitional care period, feedback loops involving patient communication back to the clinic, and clinic communication back to the patient, are critical to helping patients remain stable. For example, consider monitoring weights daily after hospital discharge. A patient at home can report increasing weights to a provider, who can then recommend an increased dose of diuretic. The patient can complete the feedback loop by taking the extra medication and monitoring the return of weight back to normal.

While daily weight monitoring is a simple process improvement that relies on the principle of establishing feedback loops, many other strategies exist. One commonly employed tool is the postdischarge telephone follow-up call, which is often coupled with other interventions in a comprehensive care bundle.⁸ During the telephone call, several process-of-care defects can be corrected, including missing medications or missing information on appointment times.

Beyond the telephone, newer technologies show promise for helping develop feedback loops for patients at home. One such technology is telemonitoring, whereby physiologic information such as weight, heart rate, and blood pressure is collected and sent back to a monitoring center. While the principle holds promise, several studies have not demonstrated significantly different outcomes as compared to usual care.^13,29 Another promising technology is the CardioMEMS device (Abbott, Inc., Atlanta, GA), which can remotely transmit the pulmonary artery pressure, a physiologic signal which correlates with volume overload. There is now strong evidence supporting the efficacy of pulmonary artery pressure–guided heart failure management.^30,31

Finally, home visits can be an efficient way to communicate symptoms, enable clinical assessment, and provide recommendations. One program that implemented home visits, 24-hour nurses available by call, and telephone follow-up showed a statistically significant reduction in readmissions.³² Furthermore, a meta-analysis of randomized controlled trials comparing home health to usual care showed decreased readmissions and mortality.³³ The efficacy may be in strengthening the feedback loop—home care improves compliance with weight monitoring, fluid restriction, and medications.³⁴ These studies provide a strong rationale for the benefits of home health in stabilizing heart failure patients postdischarge. Indeed, nurse home visits were 1 of the 2 process interventions in a Cochrane review of randomized controlled trials that were shown to statistically significantly decrease readmissions and mortality.⁹ These data underscore the importance of feedback loops for helping ensure patients are clinically stable.

Postdischarge Follow-Up Clinic Visit

The first clinic appointment postdischarge is an important check-in to help advance patient care. Several key tasks can be achieved during the postdischarge visit. First, the patient can be clinically stabilized by adjusting diuretic therapy. If the patient is clinically stable, GDMT can be uptitrated. Second, education around symptoms, medications, diet, and exercise can be reinforced. Finally, clinicians can help connect patients to other members of the multidisciplinary care team, including specialist care, home health, or cardiac rehabilitation.

Achieving 7-day follow-up visits after discharge has been a point of emphasis in national guidelines.²⁴ The ACC promotes a “See You in 7” challenge, advising that all patients discharged with a diagnosis of heart failure have a follow-up appointment within 7 days. Yet based on the latest available data, arrival rates to the postdischarge clinic are dismal, hovering around 30%.³⁵ In a multicenter observational study of hospitals participating in the “See You in 7” collaborative, hospitals were able to increase their 7-day follow-up appointment rates by 2% to 3%, and also noted an absolute decrease in readmission rates by 1% to 2%.³⁶ We have demonstrated, using a mathematical approach called queuing theory, that discharge appointment wait times and clinic access can be significantly improved by providing a modest capacity buffer to clinic availability.³⁷ Those interested in applying this model to their own clinical practice may do so with a free online calculator at http://hfresearch.org.

Another important aspect of postdischarge follow-up is appropriate management of the comorbidity burden, which, as noted, is often significant in patients hospitalized with heart failure.³⁸ For instance, in recent cohorts of hospitalized heart failure patients, the incidence of hypertension was 78%, coronary artery disease was more than 50%, atrial fibrillation was more than 40%, and diabetes was nearly 40%.³⁹ Given this burden of comorbidity, it is not surprising that only 35% of readmissions after an index heart failure hospitalization are for recurrent heart failure.⁴⁰ Coordinating care among primary care physicians and relevant subspecialists is thus essential. Phone calls and secure electronic messages are very helpful in achieving this. There is increasing interest in more nimble care models, such as the patient-centered specialty practice⁴¹ or the dyspnea clinic, to help bring coordinated resources to the patient.⁴²

Process of Process Improvement: Our Experiences

The previous sections outline a series of potential process improvements clinical teams and health systems can implement to impact heart failure readmissions. A plan on paper, however, does not equal a plan in actuality. How does one go about implementing these changes? We offer our local experience starting a heart failure transitional care program as a case study, then draw lessons learned as a set of practical tips for local teams to employ. What we hope to highlight is that there is a large difference between a completed process for transitional care of heart failure patients, and the process of developing that process itself. The former is the hardware, the latter is the software. The latter does not typically get highlighted, but it is absolutely critical to unlocking the capabilities of a team and the institution.

In 2015, Northwestern Memorial Hospital adopted a novel payment arrangement from the Center for Medicare and Medicaid Services for Medicare patients being discharged from the hospital with heart failure. Known as Bundled Payments for Care Improvement,⁴³ this bundled payment model incentivized Northwestern Memorial Hospital charge, principally by reducing hospital readmissions and by collaborating with skilled nursing facilities to control length of stay.

We approached this problem by drawing on the available literature,^44,45 and by first creating a schematic of our high-level approach, which comprised 3 major elements (Figure 2): identification of hospitalized heart failure patients, delivery of a care bundle to hospitalized heart failure patients in hospital, and coordinating postdischarge care, centered on a telephone call and a postdischarge visit.

We then proceeded by building out, in stepwise fashion, each component of our value chain, using Agile techniques as a guiding principle.⁴⁶ Agile, a productivity and process improvement mindset with roots in software development, emphasizes tackling 1 problem at a time, building out new features sequentially and completely, recognizing that the end user does not derive value from a program until new functionality is available for use. Rather than wholesale monolithic change, Agile emphasizes rapid iteration, prototyping, and discarding innovations not found to be helpful. The notion is to stand up new, incremental features rapidly, with each incremental improvement delivering value and helping to accelerate overall change.

Our experience building a robust way to identify heart failure cases is a good example of Agile process improvement in practice. At our hospital, identification of patients with heart failure was a challenge because more than half of heart failure patients are admitted to noncardiology floors. We developed a simple electronic health record query to detect heart failure patients, relying on parameters such as administration of intravenous diuretic or levels of BNP exceeding 100 ng/dL. We deployed this query, finding very high sensitivity for detection of heart failure patients.¹⁴ Patients found to have heart failure were then populated into a list in the electronic health record, which made patients’ heart failure status visible to all members of the health care team. Using this list, we were able to automate several processes necessary for heart failure care. For example, the list made it possible for cardiologists to know if there was a patient who perhaps needed cardiology consultation. Nurse navigators could know which patients needed heart failure education without having to be actively consulted by the admitting team. The same nurse navigators could then know upon discharge which patients needed a follow-up telephone call at 48 hours.

This list of heart failure patients was the end product, which was built through prototyping and iteration. For example, with our initial BNP cutoff of 300 ng/dL, we recognized we were missing several cases, and lowered the cutoff for the screener to 100 ng/dL. When we were satisfied this process was working well, we moved on to the next problem to tackle, avoiding trying to work on too many things at once. By doing so, we were able to focus our process improvement resources on 1 problem at a time, building up a suite of interventions. For our hospital, we settled on a bundle of interventions, captured by the mnemonic HEART:

Heart doctor sees patient in the hospital

Education about heart failure in the hospital

After-visit summary with 7-day appointment printed

Reach out to the patient by telephone within 72 hours

Treat the patient in clinic by the 7-day visit

Conclusion

We would like to emphasize that the elements of our heart failure readmissions interventions were not all put in place at once. This was an iterative process that proceeded in a stepwise fashion, with each step improving the care of our patients. We learned a number of lessons from our experience. First, we would advise that teams not try to do everything. One program simply cannot implement all possible readmission reduction interventions, and certainly not all at once. Trade-offs should be made, and interventions more likely to succeed in the local environment should be prioritized. In addition, interventions that do not fit and do not create synergy with the local practice environment should not be pursued.

Second, we would advise teams to start small, tackling a known problem in heart failure transitions of care first. This initial intuition is often right. An example might be improving 7-day appointments upon discharge. Starting with a problem that can be tackled builds process improvement muscle and improves team morale. Third, we would advise teams to consistently iterate on designs, tweaking and improving performance. Complex organizations always evolve; processes that work 1 year may fail the next because another element of the organization may have changed.

Finally, the framework presented in Figure 1 may be helpful in guiding how to structure interventions. Considering interventions to be delivered in the hospital, interventions to be delivered in the clinic, and how to set up feedback loops to support patients as outpatients help develop a comprehensive heart failure readmissions reduction program.

Corresponding author: R. Kannan Mutharasan, MD, Northwestern University Feinberg School of Medicine, 676 North Saint Clair St., Arkes Pavilion, Suite 7-038, Chicago, IL 60611;[email protected].

Financial disclosures: None.

From the Department of Medicine, Division of Cardiology, Northwestern University Feinberg School of Medicine, Chicago, IL.

Abstract

Objective: To review selected process-of-care interventions that can be applied both during the hospitalization and during the transitional care period to help address the persistent challenge of heart failure readmissions.

Methods: Review of the literature.

Results: Process-of-care interventions that can be implemented to reduce readmissions of heart failure patients include: accurately identifying heart failure patients; providing disease education; titrating guideline-directed medical therapy; ensuring discharge readiness; arranging close discharge follow-up; identifying and addressing social barriers; following up by telephone; using home health; and addressing comorbidities. Importantly, the heart failure hospitalization is an opportunity to set up outpatient success, and setting up feedback loops can aid in post-discharge monitoring.

Conclusion: We encourage teams to consider local capabilities when selecting processes to improve; begin by improving something small to build capacity and team morale, and continually iterate and reexamine processes, as health care systems are continually evolving.

Keywords: heart failure; process improvement; quality improvement; readmission; rehospitalization; transitional care.

The growing population of patients affected by heart failure continues to challenge health systems. The increasing prevalence is paralleled by the rising costs of managing heart failure, which are projected to grow from $30.7 billion in 2012 to $69.8 billion in 2030.¹ A significant portion of these costs relate to readmission after an index heart failure hospitalization. The statistics are staggering: for patients hospitalized with heart failure, approximately 15% to 20% are readmitted within 30 days.^2,3 Though recent temporal trends suggest a modest reduction in readmission rates, there is a concerning correlation with increasing mortality,³ and a recognition that readmission rate decreases may relate to subtle changes in coding-based risk adjustment.⁴ Despite these concerns, efforts to reduce readmissions after heart failure hospitalization command significant attention.

Process improvement methodologies may be helpful in reducing hospital readmissions. Various approaches have been employed, and results have been mixed. An analysis of 70 participating hospitals in the American Heart Association’s Get With the Guidelines initiative found that, while overall readmission rates declined by 1.0% over 3 years, only 1 hospital achieved a 20% reduction in readmission rates.⁵

It is notably difficult to reduce readmissions after heart failure hospitalization. One challenge is that patients with heart failure often have multiple comorbidities, and approximately 50% to 60% of 30-day readmissions after heart failure hospitalization arise from noncardiac causes.¹ Another challenge is that a significant fraction of readmissions in general—perhaps 75%—may not be avoidable.⁶

Recent excellent systematic reviews and meta-analyses provide comprehensive overviews of process improvement strategies that can be used to reduce readmissions after heart failure hospitalizations.^7-9 Yet despite this extensive knowledge, few reports discuss the process of actually implementing these changes: the process of process improvement. Here, we seek to not only highlight some of the most promising potential interventions to reduce heart failure readmissions, but also to discuss a process improvement framework to help engender success, using our experience as a case study. We schematize process improvement efforts as having several distinct phases (Figure 1): processes delivered during the hospitalization and prior to discharge; feedback loops set up to maintain clinical stability at home; and the postdischarge clinic visit as an opportunity to further stabilize the patient and advance the plan of care. The discussion of these interventions follows this organization.

During Hospitalization

The heart failure hospitalization can be used as an opportunity to set up outpatient success, with several goals to target during the index admission. One goal is identifying the root causes of the heart failure syndrome and correcting those root causes, if possible. For example, patients in whom the heart failure syndrome is secondary to valvular heart disease may benefit from transcatheter aortic valve replacement.¹⁰ Another clinical goal is decongesting the patient, which is associated with lower readmission rates.^11,12 These goals focus on the medical aspects of heart failure care. However, beyond these medical aspects, a patient must be equipped to successfully manage the disease at home.

To support medical and nonmedical interventions for hospitalized heart failure patients, a critical first step is identifying patients with heart failure. This accomplishes at least 2 objectives. First, early identification allows early initiation of interventions, such as heart failure education and social work evaluation. Early initiation of these interventions allows sufficient time during the hospitalization to make meaningful progress on these fronts. Second, early identification allows an opportunity for the delivery of cardiology specialty care, which may help with identifying and correcting root causes of the heart failure syndrome. Such access to cardiology has been shown to improve inpatient mortality and readmission rates.¹³

In smaller hospitals, identification of patients with heart failure can be as simple as reviewing overnight admissions. More advanced strategies, such as screeners based on brain natriuretic peptide (BNP) levels and administration of intravenous diuretics, can be employed.^14,15 In the near future, deep learning-based natural language processing will be applied to mine full-text data in the electronic health record to identify heart failure hospitalizations.¹⁶

In the hospital, patients can also receive education about heart failure disease management. This education is a cornerstone of reducing heart failure readmissions. A recent systematic review of nurse education interventions demonstrated reductions in readmissions, hospitalizations, and costs.¹⁷ However, the efficacy of heart failure education hinges on many other variables. For patients to adhere to water restriction and daily weights, for example, there must also be patient understanding, compliance, and accessibility to providers to recommend how to strike the fluid balance. Education is therefore necessary, but not sufficient, for setting up outpatient success.

The hospitalization also represents an important time to start or uptitrate guideline-directed medical therapy (GDMT) for heart failure. Doing so takes advantage of an important opportunity to reduce the risk of readmission and even reverse the disease process.¹⁸ Uptitration of GDMT in patients with heart failure with reduced ejection fraction is associated with a decreased risk of mortality, while discontinuation is associated with an increased risk of mortality.¹⁹ However, recent registry data indicate that intensity of GDMT is just as likely to be decreased as increased during the hospitalization.²⁰ Nevertheless, predischarge initiation of medications may be associated with higher attained doses in follow-up.²¹

Preparing for Discharge

Preparing a patient for discharge after a heart failure hospitalization involves stabilizing the medical condition as well as ensuring that the patient and caregivers have the medication, equipment, and self-care resources at home necessary to manage the condition. Several frameworks have been put forth to help care teams analyze a patient’s readiness for discharge. One is the B-PREPARED score,²² a validated instrument to discriminate among patients with regard to their readiness to discharge from the hospital. This instrument highlights the importance of several key factors that should be addressed during the discharge process, including counseling and written instructions about medications and their side effects; information about equipment needs and community resources; and information on activity levels and restrictions. Nurse education and discharge coordination can improve patients’ perception of discharge readiness,²³ although whether this discharge readiness translates into improved readmission rates appears to depend on the specific follow-up intervention design.⁹

Prior to discharge, it is important to arrange postdischarge follow-up appointments, as emphasized by the American College of Cardiology/American Heart Association (ACC/AHA) guidelines.²⁴ The use of nurse navigators can help with planning follow-up appointments. For example, the ACC Patient Navigator Program was applied in a single-center study of 120 patients randomized to the program versus usual care.²⁵ This study found a significant increase in patient education and follow-up appointments compared to usual care, and a numerical decrease in hospital readmissions, although the finding was not statistically significant.²⁵

A third critical component of preparing for discharge is identifying and addressing social barriers to care. In a study of patients stratified by household income, patients in the lowest income quartile had a higher readmission rate than patients in the highest income quartile.²⁶ Poverty also correlates with heart failure mortality.²⁷ Social factors play an important role in many aspects of patients’ ability to manage their health, including self-care, medication adherence, and ability to follow-up. Identifying these social factors prior to discharge is the first step to addressing them. While few studies specifically address the role of social workers in the management of heart failure care, the general medical literature suggests that social workers embedded in transitional care teams can augment readmission reduction efforts.²⁸

After Discharge

Patients recently discharged from the hospital who have not yet attended their postdischarge appointment are in an incredibly vulnerable phase of care. Patients who are discharged from the hospital may not yet be connected with outpatient care. During this initial transitional care period, feedback loops involving patient communication back to the clinic, and clinic communication back to the patient, are critical to helping patients remain stable. For example, consider monitoring weights daily after hospital discharge. A patient at home can report increasing weights to a provider, who can then recommend an increased dose of diuretic. The patient can complete the feedback loop by taking the extra medication and monitoring the return of weight back to normal.

While daily weight monitoring is a simple process improvement that relies on the principle of establishing feedback loops, many other strategies exist. One commonly employed tool is the postdischarge telephone follow-up call, which is often coupled with other interventions in a comprehensive care bundle.⁸ During the telephone call, several process-of-care defects can be corrected, including missing medications or missing information on appointment times.

Beyond the telephone, newer technologies show promise for helping develop feedback loops for patients at home. One such technology is telemonitoring, whereby physiologic information such as weight, heart rate, and blood pressure is collected and sent back to a monitoring center. While the principle holds promise, several studies have not demonstrated significantly different outcomes as compared to usual care.^13,29 Another promising technology is the CardioMEMS device (Abbott, Inc., Atlanta, GA), which can remotely transmit the pulmonary artery pressure, a physiologic signal which correlates with volume overload. There is now strong evidence supporting the efficacy of pulmonary artery pressure–guided heart failure management.^30,31

Finally, home visits can be an efficient way to communicate symptoms, enable clinical assessment, and provide recommendations. One program that implemented home visits, 24-hour nurses available by call, and telephone follow-up showed a statistically significant reduction in readmissions.³² Furthermore, a meta-analysis of randomized controlled trials comparing home health to usual care showed decreased readmissions and mortality.³³ The efficacy may be in strengthening the feedback loop—home care improves compliance with weight monitoring, fluid restriction, and medications.³⁴ These studies provide a strong rationale for the benefits of home health in stabilizing heart failure patients postdischarge. Indeed, nurse home visits were 1 of the 2 process interventions in a Cochrane review of randomized controlled trials that were shown to statistically significantly decrease readmissions and mortality.⁹ These data underscore the importance of feedback loops for helping ensure patients are clinically stable.

Postdischarge Follow-Up Clinic Visit

The first clinic appointment postdischarge is an important check-in to help advance patient care. Several key tasks can be achieved during the postdischarge visit. First, the patient can be clinically stabilized by adjusting diuretic therapy. If the patient is clinically stable, GDMT can be uptitrated. Second, education around symptoms, medications, diet, and exercise can be reinforced. Finally, clinicians can help connect patients to other members of the multidisciplinary care team, including specialist care, home health, or cardiac rehabilitation.

Achieving 7-day follow-up visits after discharge has been a point of emphasis in national guidelines.²⁴ The ACC promotes a “See You in 7” challenge, advising that all patients discharged with a diagnosis of heart failure have a follow-up appointment within 7 days. Yet based on the latest available data, arrival rates to the postdischarge clinic are dismal, hovering around 30%.³⁵ In a multicenter observational study of hospitals participating in the “See You in 7” collaborative, hospitals were able to increase their 7-day follow-up appointment rates by 2% to 3%, and also noted an absolute decrease in readmission rates by 1% to 2%.³⁶ We have demonstrated, using a mathematical approach called queuing theory, that discharge appointment wait times and clinic access can be significantly improved by providing a modest capacity buffer to clinic availability.³⁷ Those interested in applying this model to their own clinical practice may do so with a free online calculator at http://hfresearch.org.

Another important aspect of postdischarge follow-up is appropriate management of the comorbidity burden, which, as noted, is often significant in patients hospitalized with heart failure.³⁸ For instance, in recent cohorts of hospitalized heart failure patients, the incidence of hypertension was 78%, coronary artery disease was more than 50%, atrial fibrillation was more than 40%, and diabetes was nearly 40%.³⁹ Given this burden of comorbidity, it is not surprising that only 35% of readmissions after an index heart failure hospitalization are for recurrent heart failure.⁴⁰ Coordinating care among primary care physicians and relevant subspecialists is thus essential. Phone calls and secure electronic messages are very helpful in achieving this. There is increasing interest in more nimble care models, such as the patient-centered specialty practice⁴¹ or the dyspnea clinic, to help bring coordinated resources to the patient.⁴²

Process of Process Improvement: Our Experiences

The previous sections outline a series of potential process improvements clinical teams and health systems can implement to impact heart failure readmissions. A plan on paper, however, does not equal a plan in actuality. How does one go about implementing these changes? We offer our local experience starting a heart failure transitional care program as a case study, then draw lessons learned as a set of practical tips for local teams to employ. What we hope to highlight is that there is a large difference between a completed process for transitional care of heart failure patients, and the process of developing that process itself. The former is the hardware, the latter is the software. The latter does not typically get highlighted, but it is absolutely critical to unlocking the capabilities of a team and the institution.

In 2015, Northwestern Memorial Hospital adopted a novel payment arrangement from the Center for Medicare and Medicaid Services for Medicare patients being discharged from the hospital with heart failure. Known as Bundled Payments for Care Improvement,⁴³ this bundled payment model incentivized Northwestern Memorial Hospital charge, principally by reducing hospital readmissions and by collaborating with skilled nursing facilities to control length of stay.

We approached this problem by drawing on the available literature,^44,45 and by first creating a schematic of our high-level approach, which comprised 3 major elements (Figure 2): identification of hospitalized heart failure patients, delivery of a care bundle to hospitalized heart failure patients in hospital, and coordinating postdischarge care, centered on a telephone call and a postdischarge visit.

We then proceeded by building out, in stepwise fashion, each component of our value chain, using Agile techniques as a guiding principle.⁴⁶ Agile, a productivity and process improvement mindset with roots in software development, emphasizes tackling 1 problem at a time, building out new features sequentially and completely, recognizing that the end user does not derive value from a program until new functionality is available for use. Rather than wholesale monolithic change, Agile emphasizes rapid iteration, prototyping, and discarding innovations not found to be helpful. The notion is to stand up new, incremental features rapidly, with each incremental improvement delivering value and helping to accelerate overall change.

Our experience building a robust way to identify heart failure cases is a good example of Agile process improvement in practice. At our hospital, identification of patients with heart failure was a challenge because more than half of heart failure patients are admitted to noncardiology floors. We developed a simple electronic health record query to detect heart failure patients, relying on parameters such as administration of intravenous diuretic or levels of BNP exceeding 100 ng/dL. We deployed this query, finding very high sensitivity for detection of heart failure patients.¹⁴ Patients found to have heart failure were then populated into a list in the electronic health record, which made patients’ heart failure status visible to all members of the health care team. Using this list, we were able to automate several processes necessary for heart failure care. For example, the list made it possible for cardiologists to know if there was a patient who perhaps needed cardiology consultation. Nurse navigators could know which patients needed heart failure education without having to be actively consulted by the admitting team. The same nurse navigators could then know upon discharge which patients needed a follow-up telephone call at 48 hours.

This list of heart failure patients was the end product, which was built through prototyping and iteration. For example, with our initial BNP cutoff of 300 ng/dL, we recognized we were missing several cases, and lowered the cutoff for the screener to 100 ng/dL. When we were satisfied this process was working well, we moved on to the next problem to tackle, avoiding trying to work on too many things at once. By doing so, we were able to focus our process improvement resources on 1 problem at a time, building up a suite of interventions. For our hospital, we settled on a bundle of interventions, captured by the mnemonic HEART:

Heart doctor sees patient in the hospital

Education about heart failure in the hospital

After-visit summary with 7-day appointment printed

Reach out to the patient by telephone within 72 hours

Treat the patient in clinic by the 7-day visit

Conclusion

We would like to emphasize that the elements of our heart failure readmissions interventions were not all put in place at once. This was an iterative process that proceeded in a stepwise fashion, with each step improving the care of our patients. We learned a number of lessons from our experience. First, we would advise that teams not try to do everything. One program simply cannot implement all possible readmission reduction interventions, and certainly not all at once. Trade-offs should be made, and interventions more likely to succeed in the local environment should be prioritized. In addition, interventions that do not fit and do not create synergy with the local practice environment should not be pursued.

Second, we would advise teams to start small, tackling a known problem in heart failure transitions of care first. This initial intuition is often right. An example might be improving 7-day appointments upon discharge. Starting with a problem that can be tackled builds process improvement muscle and improves team morale. Third, we would advise teams to consistently iterate on designs, tweaking and improving performance. Complex organizations always evolve; processes that work 1 year may fail the next because another element of the organization may have changed.

Finally, the framework presented in Figure 1 may be helpful in guiding how to structure interventions. Considering interventions to be delivered in the hospital, interventions to be delivered in the clinic, and how to set up feedback loops to support patients as outpatients help develop a comprehensive heart failure readmissions reduction program.

Corresponding author: R. Kannan Mutharasan, MD, Northwestern University Feinberg School of Medicine, 676 North Saint Clair St., Arkes Pavilion, Suite 7-038, Chicago, IL 60611;[email protected].

Financial disclosures: None.