As the field of hospital medicine grows, some hospitalists are gravitating toward subspecialty services. In recent years we’ve witnessed a proliferation of ‘ists’: There are now surgicalists, laborists, psychiatric hospitalists—even hepa-hospitalists.

The numbers of “hyphenate hospitalists” are not tracked by SHM, but the subspecialization trend highlights raises questions about hospital medicine’s evolution. Among the issues:

• What does this growth of hospitalist subspecialists foreshadow about the strength of the hospitalist movement?
• Does subspecialization always convey positive changes for the hospitalist?
• Do physicians risk trade-offs when their hospital medicine practices are rooted solely in one subspecialty?
• What about retaining the opportunity to see and treat a variety of patients and conditions—presumably one of the initial attractions of a career in internal medicine and family medicine?

The founder of hospital medicine, a noted pediatric hospitalist, the chair of the SHM’s membership committee, and a former hepa-hospitalist recently shared their experiences and views on these issues.

Success Spreads

Hospital medicine pioneer Robert M. Wachter, MD, has observed at his and other hospitals the increasing dependence on hospitalists’ services.

“Hospitalists have traditionally done more than just take care of medical patients,” says Dr. Wachter, professor and chief of the division of hospital medicine, associate chairman, department of medicine, chief of the medical service at the University of California San Francisco (UCSF), and author of the upcoming blog “Wachter’s World.” “They’ve always done medical consultations and helped to take care of sick patients with surgical, gynecological, and psychiatric issues.” But now, he says, “The demand for hospitalist services is almost limitless.”

At UCSF, he reports, hospitalists now manage the medical problems of patients on the complex heart failure service, the bone marrow transplant service, and the neurosurgical and orthopedic services. Dr. Wachter views the trend of using hospitalists in a variety of subspecialty services as “one of the most exciting developments for the field—it is taking the field to a whole new level of importance and growth.” That’s because it signals recognition that the concept of hospital medicine has value “for virtually every patient sick enough to be in the building,” he says.

Ambiguity of Terms

Not only are hospitalists increasingly present in subspecialty services, but some specialist services are reorganizing according to the hospitalist model. This may create complexities regarding hospital medicine’s core identity, according to Dr. Wachter.

For instance, at UCSF, there are generalist surgeons who have organized a hospitalist service, providing on-call responsiveness, triage for specialized surgical problems, and a breadth of care and coordination typical of the hospital medicine model. Separately, there are also internal medicine hospitalists who serve on the surgery service. “I think there is going to be some ambiguity about roles until we clean up the language,” remarks Dr. Wachter. For instance: “Is the hospitalist on the surgery service still a generalist who takes on the role of subspecialist by caring for a more specialized population? And, what do you call the specialist surgeon who takes on a more hospitalist role?”

Shaun Frost, MD, FACP, chair of SHM’s Membership Committee and regional medical director for Cogent Healthcare in St. Paul, Minn., considers whether the inclusion of various subspecialists fits with the SHM’s definition of hospitalists. “If you’re looking strictly at ‘definition,’ SHM considers a hospitalist to be a physician whose primary professional focus is the general medical care of hospitalized patients,” he says. “As hospitalist subspecialists are likely engaging in the management of hypertension, diabetes, chronic lung and cardiac disease, etc., I see no reason to believe that they would not fit the definition of a hospitalist.”

Best Use of Skills?

Although Drs. Wachter and Frost see inclusion of hospitalists on subspecialty services as a positive trend, others warn that hospitalists should be wary about the reasons for their enlistment.

Lauren M. Friedly, MD, a hospitalist at Marin General Hospital in Greenbrae, Calif., believes subspecialty hospitalist jobs are “ultimately untenable for solid, well-trained, dedicated hospitalists.” She developed this view after a frustrating two years on a liver transplant service—where she found she “wasn’t able to practice medicine in a way that was comfortable.”

Dr. Friedly explains that she chose to be a hospitalist because of her experience as a medical student at UCSF and as a resident at California Pacific Medical Center—watching and learning from the originators of the movement, such as Dr. Wachter and Masa Yukimoto, MD, former chief resident. All the reasons she chose hospital medicine—the pace, acuity of patients, ability to revisit patients and ruminate about their problems, and the opportunity to improve the quality of a patient’s in-hospital care by adhering to a “first do no harm” philosophy—were stymied when she became a hepa-hospitalist.

“The problems that can potentially exist in any subspecialty hospitalist group are magnified a hundredfold in a liver transplant program,” she explains. “There were philosophical differences in our approach to medical care of inpatients, and the hospitalists were, in some ways, considered the bottom of the food chain.

“Because of our position relative to the hepatologists and transplant surgeons, we were not provided the autonomy nor the resources with which to accomplish any of these things. For example, decreasing length of stay by discharging patients efficiently, which to a well-trained hospitalist may mean less risk of exposure to nosocomial infections and iatrogenic complications, was not necessarily valued by the transplant surgeons and hepatologists. Less is often more for a well-trained hospitalist, but this sentiment is in direct conflict with the maximalist approach used in transplant medicine.”

Satisfying in the Long Run?

Perhaps the most important question about subspecialization for the hospitalist is whether joining a subspecialist service is a good fit. Dr. Frost believes it’s important to consider whether a career in a subspecialty service would be satisfying in the long run. “For many of us,” he explains, “one main reason we chose to pursue careers in general internal medicine, general pediatrics, or family practice is that we enjoy variety. Personally, for example, I know that I would quickly become bored with solely focusing on one organ system or one special patient population. Therefore, I believe that there is probably a limited group of folks who would enjoy exclusively restricting the scope of their hospital medicine practice to a specific subspecialty area.”

Still, Dr. Friedly admits she will miss the challenges of dealing with complex pathology. On the liver transplant service, for instance, she cared for patients with complex gastrointestinal issues, including cancers. But, due to her time on that specialized service, she now finds herself having to refresh her more general internal medicine skills.

The Pediatrics Picture

The hospitalist model is increasingly common in pediatrics, where costs per patient and length of stay have been lowered when using the hospital medicine model to restructure academic pediatric inpatient services, and hospitalists have contributed to improved survival in pediatric intensive care units.1-2

In part because of lower patient volumes on pediatric hospitalist services, the trend toward subspecialization is not as evident in pediatric hospital medicine as it is in adult hospital medicine, according to Sanford M. Melzer, MD, senior vice president of strategic planning and business development at Seattle Children’s Hospital and Regional Medical Center, and a member of the American Academy of Pediatrics’ Committee for Hospital Care. However, with shortages of physicians in key specialty areas, that may be changing.

Dr. Melzer, who has been a clinical pediatric hospitalist for 20 years and has published research on the financial aspects of pediatric hospitalist programs, reports that his service is beginning to field requests for hospitalists from specialty programs to provide staffing—just as adult hospital medicine programs have noticed.3 For example, the oncology service at Seattle Children’s Hospital has started to explore using pediatric hospitalists to help manage its pediatric cancer unit, which treats 225 new diagnoses of childhood cancer annually.

In part because oncology treatment at Seattle Children’s tends to be “heavily protocolized,” Dr. Melzer believes pediatric hospitalists can provide quality inpatient care as inpatient generalists, in terms of palliative care and symptom management, if included on that service.

The other area in which pediatric hospitalists may provide “specialty” care is in neonatology, where shortages of specialists or costs of coverage result in hospitalists covering the delivery room or the neonatal intensive care unit (NICU).

Pediatric hospitalists, as generalists, typically provide care for many different types of illnesses and conditions. “This is one of the attractive features of the job for pediatricians choosing this career track,” Dr. Melzer says. “An increasing degree of specialization may make these positions somewhat less interesting, and may highlight discrepancies between hospitalist and specialist salaries.” On the other hand, he says, “Continued shortages in specialty areas in pediatrics will continue to drive the trends toward increased deployment of generalists in these services.”

One model employed in children’s hospitals is to utilize more physician extenders, such as nurse practitioners and physician assistants, to provide the needed coverage. How pediatric hospitalists will fit into this evolving care model is not clear, notes Dr. Melzer.

Future Configurations

To avoid the possible pitfalls hospitalists can encounter with subspecialist services will require innovative solutions, Dr. Friedly believes.

“Ultimately, I think the only way that it will be sustainable [as a long-term career choice] for any individual hospitalist to take a position within a subspecialty program will be to help create or to be part of the initial vision,” Dr. Friedly says. “Or, an already-established hospitalist group could approach the subspecialist and ask, ‘How can we help you deliver the highest quality standard of care to your patients while they’re here in the hospital?’ ”

As of July 1, the UCSF neurosurgery service has embraced the addition of core internal medicine hospitalists who help to manage the medical problems of the [typical census of] 60 neurosurgery patients. So far, Dr. Wachter reports, the hospitalists are enjoying their stint on the neurosurgery service. “They’re learning a lot because these patients have very unusual and specialized problems,’’ he says. “It only took three minutes for them to realize that they were making a difference because some of the patients are very sick and have many medical problems. The surgeons don’t focus on or keep up with medical management, and even if they did, they are in the OR all day long.”

Rotating hospitalists to specialty and subspecialty services for 25% of their time in the hospital, as the UCSF Neurosurgery hospitalists do, might be one way to preserve the traditional general medical model to which many internal medicine hospitalists still gravitate. “This could be an interesting, specialized niche practice, but would not be the bulk of what they do,” says Dr. Wachter.

Dr. Frost agrees that the key to addressing the challenge of subspecialization lies in building this type of work into the context of a larger hospital medicine program. “Rotating all members of a hospital medicine group through a subspecialty experience for a portion of their overall time may be the way to go,” he notes.

Dr. Friedly cautions that certain subspecialist services, such as liver transplant, may not embrace the multidisciplinary hospital medicine model, so it remains to be seen if the effort can evolve to be truly collaborative. Her advice to younger residents just entering hospital medicine? “Hospitalist medicine has unlimited possibilities as a career choice, especially if you enjoyed being an internal medicine resident. Be careful, however, to avoid a setting where you risk losing your hard-earned skills while also being treated like a ‘perma-resident.’ Starting out in a more traditional hospitalist program to learn solid hospitalist ‘tricks of the trade,’ then transitioning to a subspecialty program where you can offer your skills, rather than the other way around, may be the more sustainable, long-term option.” TH

Gretchen Henkel is a frequent contributor to The Hospitalist.

References

1. Ogershok PR, Li X, Palmer HC, et al. Restructuring an academic pediatric inpatient service using concepts developed by hospitalists.” Clin Pediatr. 2001 Dec.;40(12): 653-660.
2. Tenner PA, Dibrell H, Taylor RP. Improved survival with hospitalists in a pediatric intensive care unit.” Crit Care Med. 2003 Mar;31(3):847-852.
3. Melzer SM, Molteni, RA, Marcuse EK, et al. Characteristics and financial performance of a pediatric faculty inpatient attending service: a resource-based relative value scale analysis. Pediatrics. 2001 Jul;108(1);79-84.

As the field of hospital medicine grows, some hospitalists are gravitating toward subspecialty services. In recent years we’ve witnessed a proliferation of ‘ists’: There are now surgicalists, laborists, psychiatric hospitalists—even hepa-hospitalists.

The numbers of “hyphenate hospitalists” are not tracked by SHM, but the subspecialization trend highlights raises questions about hospital medicine’s evolution. Among the issues:

• What does this growth of hospitalist subspecialists foreshadow about the strength of the hospitalist movement?
• Does subspecialization always convey positive changes for the hospitalist?
• Do physicians risk trade-offs when their hospital medicine practices are rooted solely in one subspecialty?
• What about retaining the opportunity to see and treat a variety of patients and conditions—presumably one of the initial attractions of a career in internal medicine and family medicine?

The founder of hospital medicine, a noted pediatric hospitalist, the chair of the SHM’s membership committee, and a former hepa-hospitalist recently shared their experiences and views on these issues.

Success Spreads

Hospital medicine pioneer Robert M. Wachter, MD, has observed at his and other hospitals the increasing dependence on hospitalists’ services.

“Hospitalists have traditionally done more than just take care of medical patients,” says Dr. Wachter, professor and chief of the division of hospital medicine, associate chairman, department of medicine, chief of the medical service at the University of California San Francisco (UCSF), and author of the upcoming blog “Wachter’s World.” “They’ve always done medical consultations and helped to take care of sick patients with surgical, gynecological, and psychiatric issues.” But now, he says, “The demand for hospitalist services is almost limitless.”

At UCSF, he reports, hospitalists now manage the medical problems of patients on the complex heart failure service, the bone marrow transplant service, and the neurosurgical and orthopedic services. Dr. Wachter views the trend of using hospitalists in a variety of subspecialty services as “one of the most exciting developments for the field—it is taking the field to a whole new level of importance and growth.” That’s because it signals recognition that the concept of hospital medicine has value “for virtually every patient sick enough to be in the building,” he says.

Ambiguity of Terms

Not only are hospitalists increasingly present in subspecialty services, but some specialist services are reorganizing according to the hospitalist model. This may create complexities regarding hospital medicine’s core identity, according to Dr. Wachter.

For instance, at UCSF, there are generalist surgeons who have organized a hospitalist service, providing on-call responsiveness, triage for specialized surgical problems, and a breadth of care and coordination typical of the hospital medicine model. Separately, there are also internal medicine hospitalists who serve on the surgery service. “I think there is going to be some ambiguity about roles until we clean up the language,” remarks Dr. Wachter. For instance: “Is the hospitalist on the surgery service still a generalist who takes on the role of subspecialist by caring for a more specialized population? And, what do you call the specialist surgeon who takes on a more hospitalist role?”

Shaun Frost, MD, FACP, chair of SHM’s Membership Committee and regional medical director for Cogent Healthcare in St. Paul, Minn., considers whether the inclusion of various subspecialists fits with the SHM’s definition of hospitalists. “If you’re looking strictly at ‘definition,’ SHM considers a hospitalist to be a physician whose primary professional focus is the general medical care of hospitalized patients,” he says. “As hospitalist subspecialists are likely engaging in the management of hypertension, diabetes, chronic lung and cardiac disease, etc., I see no reason to believe that they would not fit the definition of a hospitalist.”

Best Use of Skills?

Although Drs. Wachter and Frost see inclusion of hospitalists on subspecialty services as a positive trend, others warn that hospitalists should be wary about the reasons for their enlistment.

Lauren M. Friedly, MD, a hospitalist at Marin General Hospital in Greenbrae, Calif., believes subspecialty hospitalist jobs are “ultimately untenable for solid, well-trained, dedicated hospitalists.” She developed this view after a frustrating two years on a liver transplant service—where she found she “wasn’t able to practice medicine in a way that was comfortable.”

Dr. Friedly explains that she chose to be a hospitalist because of her experience as a medical student at UCSF and as a resident at California Pacific Medical Center—watching and learning from the originators of the movement, such as Dr. Wachter and Masa Yukimoto, MD, former chief resident. All the reasons she chose hospital medicine—the pace, acuity of patients, ability to revisit patients and ruminate about their problems, and the opportunity to improve the quality of a patient’s in-hospital care by adhering to a “first do no harm” philosophy—were stymied when she became a hepa-hospitalist.

“The problems that can potentially exist in any subspecialty hospitalist group are magnified a hundredfold in a liver transplant program,” she explains. “There were philosophical differences in our approach to medical care of inpatients, and the hospitalists were, in some ways, considered the bottom of the food chain.

“Because of our position relative to the hepatologists and transplant surgeons, we were not provided the autonomy nor the resources with which to accomplish any of these things. For example, decreasing length of stay by discharging patients efficiently, which to a well-trained hospitalist may mean less risk of exposure to nosocomial infections and iatrogenic complications, was not necessarily valued by the transplant surgeons and hepatologists. Less is often more for a well-trained hospitalist, but this sentiment is in direct conflict with the maximalist approach used in transplant medicine.”

Satisfying in the Long Run?

Perhaps the most important question about subspecialization for the hospitalist is whether joining a subspecialist service is a good fit. Dr. Frost believes it’s important to consider whether a career in a subspecialty service would be satisfying in the long run. “For many of us,” he explains, “one main reason we chose to pursue careers in general internal medicine, general pediatrics, or family practice is that we enjoy variety. Personally, for example, I know that I would quickly become bored with solely focusing on one organ system or one special patient population. Therefore, I believe that there is probably a limited group of folks who would enjoy exclusively restricting the scope of their hospital medicine practice to a specific subspecialty area.”

Still, Dr. Friedly admits she will miss the challenges of dealing with complex pathology. On the liver transplant service, for instance, she cared for patients with complex gastrointestinal issues, including cancers. But, due to her time on that specialized service, she now finds herself having to refresh her more general internal medicine skills.

The Pediatrics Picture

The hospitalist model is increasingly common in pediatrics, where costs per patient and length of stay have been lowered when using the hospital medicine model to restructure academic pediatric inpatient services, and hospitalists have contributed to improved survival in pediatric intensive care units.1-2

In part because of lower patient volumes on pediatric hospitalist services, the trend toward subspecialization is not as evident in pediatric hospital medicine as it is in adult hospital medicine, according to Sanford M. Melzer, MD, senior vice president of strategic planning and business development at Seattle Children’s Hospital and Regional Medical Center, and a member of the American Academy of Pediatrics’ Committee for Hospital Care. However, with shortages of physicians in key specialty areas, that may be changing.

Dr. Melzer, who has been a clinical pediatric hospitalist for 20 years and has published research on the financial aspects of pediatric hospitalist programs, reports that his service is beginning to field requests for hospitalists from specialty programs to provide staffing—just as adult hospital medicine programs have noticed.3 For example, the oncology service at Seattle Children’s Hospital has started to explore using pediatric hospitalists to help manage its pediatric cancer unit, which treats 225 new diagnoses of childhood cancer annually.

In part because oncology treatment at Seattle Children’s tends to be “heavily protocolized,” Dr. Melzer believes pediatric hospitalists can provide quality inpatient care as inpatient generalists, in terms of palliative care and symptom management, if included on that service.

The other area in which pediatric hospitalists may provide “specialty” care is in neonatology, where shortages of specialists or costs of coverage result in hospitalists covering the delivery room or the neonatal intensive care unit (NICU).

Pediatric hospitalists, as generalists, typically provide care for many different types of illnesses and conditions. “This is one of the attractive features of the job for pediatricians choosing this career track,” Dr. Melzer says. “An increasing degree of specialization may make these positions somewhat less interesting, and may highlight discrepancies between hospitalist and specialist salaries.” On the other hand, he says, “Continued shortages in specialty areas in pediatrics will continue to drive the trends toward increased deployment of generalists in these services.”

One model employed in children’s hospitals is to utilize more physician extenders, such as nurse practitioners and physician assistants, to provide the needed coverage. How pediatric hospitalists will fit into this evolving care model is not clear, notes Dr. Melzer.

Future Configurations

To avoid the possible pitfalls hospitalists can encounter with subspecialist services will require innovative solutions, Dr. Friedly believes.

“Ultimately, I think the only way that it will be sustainable [as a long-term career choice] for any individual hospitalist to take a position within a subspecialty program will be to help create or to be part of the initial vision,” Dr. Friedly says. “Or, an already-established hospitalist group could approach the subspecialist and ask, ‘How can we help you deliver the highest quality standard of care to your patients while they’re here in the hospital?’ ”

As of July 1, the UCSF neurosurgery service has embraced the addition of core internal medicine hospitalists who help to manage the medical problems of the [typical census of] 60 neurosurgery patients. So far, Dr. Wachter reports, the hospitalists are enjoying their stint on the neurosurgery service. “They’re learning a lot because these patients have very unusual and specialized problems,’’ he says. “It only took three minutes for them to realize that they were making a difference because some of the patients are very sick and have many medical problems. The surgeons don’t focus on or keep up with medical management, and even if they did, they are in the OR all day long.”

Rotating hospitalists to specialty and subspecialty services for 25% of their time in the hospital, as the UCSF Neurosurgery hospitalists do, might be one way to preserve the traditional general medical model to which many internal medicine hospitalists still gravitate. “This could be an interesting, specialized niche practice, but would not be the bulk of what they do,” says Dr. Wachter.

Dr. Frost agrees that the key to addressing the challenge of subspecialization lies in building this type of work into the context of a larger hospital medicine program. “Rotating all members of a hospital medicine group through a subspecialty experience for a portion of their overall time may be the way to go,” he notes.

Dr. Friedly cautions that certain subspecialist services, such as liver transplant, may not embrace the multidisciplinary hospital medicine model, so it remains to be seen if the effort can evolve to be truly collaborative. Her advice to younger residents just entering hospital medicine? “Hospitalist medicine has unlimited possibilities as a career choice, especially if you enjoyed being an internal medicine resident. Be careful, however, to avoid a setting where you risk losing your hard-earned skills while also being treated like a ‘perma-resident.’ Starting out in a more traditional hospitalist program to learn solid hospitalist ‘tricks of the trade,’ then transitioning to a subspecialty program where you can offer your skills, rather than the other way around, may be the more sustainable, long-term option.” TH

Gretchen Henkel is a frequent contributor to The Hospitalist.

References

1. Ogershok PR, Li X, Palmer HC, et al. Restructuring an academic pediatric inpatient service using concepts developed by hospitalists.” Clin Pediatr. 2001 Dec.;40(12): 653-660.
2. Tenner PA, Dibrell H, Taylor RP. Improved survival with hospitalists in a pediatric intensive care unit.” Crit Care Med. 2003 Mar;31(3):847-852.
3. Melzer SM, Molteni, RA, Marcuse EK, et al. Characteristics and financial performance of a pediatric faculty inpatient attending service: a resource-based relative value scale analysis. Pediatrics. 2001 Jul;108(1);79-84.

As the field of hospital medicine grows, some hospitalists are gravitating toward subspecialty services. In recent years we’ve witnessed a proliferation of ‘ists’: There are now surgicalists, laborists, psychiatric hospitalists—even hepa-hospitalists.

The numbers of “hyphenate hospitalists” are not tracked by SHM, but the subspecialization trend highlights raises questions about hospital medicine’s evolution. Among the issues:

• What does this growth of hospitalist subspecialists foreshadow about the strength of the hospitalist movement?
• Does subspecialization always convey positive changes for the hospitalist?
• Do physicians risk trade-offs when their hospital medicine practices are rooted solely in one subspecialty?
• What about retaining the opportunity to see and treat a variety of patients and conditions—presumably one of the initial attractions of a career in internal medicine and family medicine?

The founder of hospital medicine, a noted pediatric hospitalist, the chair of the SHM’s membership committee, and a former hepa-hospitalist recently shared their experiences and views on these issues.

Success Spreads

Hospital medicine pioneer Robert M. Wachter, MD, has observed at his and other hospitals the increasing dependence on hospitalists’ services.

“Hospitalists have traditionally done more than just take care of medical patients,” says Dr. Wachter, professor and chief of the division of hospital medicine, associate chairman, department of medicine, chief of the medical service at the University of California San Francisco (UCSF), and author of the upcoming blog “Wachter’s World.” “They’ve always done medical consultations and helped to take care of sick patients with surgical, gynecological, and psychiatric issues.” But now, he says, “The demand for hospitalist services is almost limitless.”

At UCSF, he reports, hospitalists now manage the medical problems of patients on the complex heart failure service, the bone marrow transplant service, and the neurosurgical and orthopedic services. Dr. Wachter views the trend of using hospitalists in a variety of subspecialty services as “one of the most exciting developments for the field—it is taking the field to a whole new level of importance and growth.” That’s because it signals recognition that the concept of hospital medicine has value “for virtually every patient sick enough to be in the building,” he says.

Ambiguity of Terms

Not only are hospitalists increasingly present in subspecialty services, but some specialist services are reorganizing according to the hospitalist model. This may create complexities regarding hospital medicine’s core identity, according to Dr. Wachter.

For instance, at UCSF, there are generalist surgeons who have organized a hospitalist service, providing on-call responsiveness, triage for specialized surgical problems, and a breadth of care and coordination typical of the hospital medicine model. Separately, there are also internal medicine hospitalists who serve on the surgery service. “I think there is going to be some ambiguity about roles until we clean up the language,” remarks Dr. Wachter. For instance: “Is the hospitalist on the surgery service still a generalist who takes on the role of subspecialist by caring for a more specialized population? And, what do you call the specialist surgeon who takes on a more hospitalist role?”

Shaun Frost, MD, FACP, chair of SHM’s Membership Committee and regional medical director for Cogent Healthcare in St. Paul, Minn., considers whether the inclusion of various subspecialists fits with the SHM’s definition of hospitalists. “If you’re looking strictly at ‘definition,’ SHM considers a hospitalist to be a physician whose primary professional focus is the general medical care of hospitalized patients,” he says. “As hospitalist subspecialists are likely engaging in the management of hypertension, diabetes, chronic lung and cardiac disease, etc., I see no reason to believe that they would not fit the definition of a hospitalist.”

Best Use of Skills?

Although Drs. Wachter and Frost see inclusion of hospitalists on subspecialty services as a positive trend, others warn that hospitalists should be wary about the reasons for their enlistment.

Lauren M. Friedly, MD, a hospitalist at Marin General Hospital in Greenbrae, Calif., believes subspecialty hospitalist jobs are “ultimately untenable for solid, well-trained, dedicated hospitalists.” She developed this view after a frustrating two years on a liver transplant service—where she found she “wasn’t able to practice medicine in a way that was comfortable.”

Dr. Friedly explains that she chose to be a hospitalist because of her experience as a medical student at UCSF and as a resident at California Pacific Medical Center—watching and learning from the originators of the movement, such as Dr. Wachter and Masa Yukimoto, MD, former chief resident. All the reasons she chose hospital medicine—the pace, acuity of patients, ability to revisit patients and ruminate about their problems, and the opportunity to improve the quality of a patient’s in-hospital care by adhering to a “first do no harm” philosophy—were stymied when she became a hepa-hospitalist.

“The problems that can potentially exist in any subspecialty hospitalist group are magnified a hundredfold in a liver transplant program,” she explains. “There were philosophical differences in our approach to medical care of inpatients, and the hospitalists were, in some ways, considered the bottom of the food chain.

“Because of our position relative to the hepatologists and transplant surgeons, we were not provided the autonomy nor the resources with which to accomplish any of these things. For example, decreasing length of stay by discharging patients efficiently, which to a well-trained hospitalist may mean less risk of exposure to nosocomial infections and iatrogenic complications, was not necessarily valued by the transplant surgeons and hepatologists. Less is often more for a well-trained hospitalist, but this sentiment is in direct conflict with the maximalist approach used in transplant medicine.”

Satisfying in the Long Run?

Perhaps the most important question about subspecialization for the hospitalist is whether joining a subspecialist service is a good fit. Dr. Frost believes it’s important to consider whether a career in a subspecialty service would be satisfying in the long run. “For many of us,” he explains, “one main reason we chose to pursue careers in general internal medicine, general pediatrics, or family practice is that we enjoy variety. Personally, for example, I know that I would quickly become bored with solely focusing on one organ system or one special patient population. Therefore, I believe that there is probably a limited group of folks who would enjoy exclusively restricting the scope of their hospital medicine practice to a specific subspecialty area.”

Still, Dr. Friedly admits she will miss the challenges of dealing with complex pathology. On the liver transplant service, for instance, she cared for patients with complex gastrointestinal issues, including cancers. But, due to her time on that specialized service, she now finds herself having to refresh her more general internal medicine skills.

The Pediatrics Picture

The hospitalist model is increasingly common in pediatrics, where costs per patient and length of stay have been lowered when using the hospital medicine model to restructure academic pediatric inpatient services, and hospitalists have contributed to improved survival in pediatric intensive care units.1-2

In part because of lower patient volumes on pediatric hospitalist services, the trend toward subspecialization is not as evident in pediatric hospital medicine as it is in adult hospital medicine, according to Sanford M. Melzer, MD, senior vice president of strategic planning and business development at Seattle Children’s Hospital and Regional Medical Center, and a member of the American Academy of Pediatrics’ Committee for Hospital Care. However, with shortages of physicians in key specialty areas, that may be changing.

Dr. Melzer, who has been a clinical pediatric hospitalist for 20 years and has published research on the financial aspects of pediatric hospitalist programs, reports that his service is beginning to field requests for hospitalists from specialty programs to provide staffing—just as adult hospital medicine programs have noticed.3 For example, the oncology service at Seattle Children’s Hospital has started to explore using pediatric hospitalists to help manage its pediatric cancer unit, which treats 225 new diagnoses of childhood cancer annually.

In part because oncology treatment at Seattle Children’s tends to be “heavily protocolized,” Dr. Melzer believes pediatric hospitalists can provide quality inpatient care as inpatient generalists, in terms of palliative care and symptom management, if included on that service.

The other area in which pediatric hospitalists may provide “specialty” care is in neonatology, where shortages of specialists or costs of coverage result in hospitalists covering the delivery room or the neonatal intensive care unit (NICU).

Pediatric hospitalists, as generalists, typically provide care for many different types of illnesses and conditions. “This is one of the attractive features of the job for pediatricians choosing this career track,” Dr. Melzer says. “An increasing degree of specialization may make these positions somewhat less interesting, and may highlight discrepancies between hospitalist and specialist salaries.” On the other hand, he says, “Continued shortages in specialty areas in pediatrics will continue to drive the trends toward increased deployment of generalists in these services.”

One model employed in children’s hospitals is to utilize more physician extenders, such as nurse practitioners and physician assistants, to provide the needed coverage. How pediatric hospitalists will fit into this evolving care model is not clear, notes Dr. Melzer.

Future Configurations

To avoid the possible pitfalls hospitalists can encounter with subspecialist services will require innovative solutions, Dr. Friedly believes.

“Ultimately, I think the only way that it will be sustainable [as a long-term career choice] for any individual hospitalist to take a position within a subspecialty program will be to help create or to be part of the initial vision,” Dr. Friedly says. “Or, an already-established hospitalist group could approach the subspecialist and ask, ‘How can we help you deliver the highest quality standard of care to your patients while they’re here in the hospital?’ ”

As of July 1, the UCSF neurosurgery service has embraced the addition of core internal medicine hospitalists who help to manage the medical problems of the [typical census of] 60 neurosurgery patients. So far, Dr. Wachter reports, the hospitalists are enjoying their stint on the neurosurgery service. “They’re learning a lot because these patients have very unusual and specialized problems,’’ he says. “It only took three minutes for them to realize that they were making a difference because some of the patients are very sick and have many medical problems. The surgeons don’t focus on or keep up with medical management, and even if they did, they are in the OR all day long.”

Rotating hospitalists to specialty and subspecialty services for 25% of their time in the hospital, as the UCSF Neurosurgery hospitalists do, might be one way to preserve the traditional general medical model to which many internal medicine hospitalists still gravitate. “This could be an interesting, specialized niche practice, but would not be the bulk of what they do,” says Dr. Wachter.

Dr. Frost agrees that the key to addressing the challenge of subspecialization lies in building this type of work into the context of a larger hospital medicine program. “Rotating all members of a hospital medicine group through a subspecialty experience for a portion of their overall time may be the way to go,” he notes.

Dr. Friedly cautions that certain subspecialist services, such as liver transplant, may not embrace the multidisciplinary hospital medicine model, so it remains to be seen if the effort can evolve to be truly collaborative. Her advice to younger residents just entering hospital medicine? “Hospitalist medicine has unlimited possibilities as a career choice, especially if you enjoyed being an internal medicine resident. Be careful, however, to avoid a setting where you risk losing your hard-earned skills while also being treated like a ‘perma-resident.’ Starting out in a more traditional hospitalist program to learn solid hospitalist ‘tricks of the trade,’ then transitioning to a subspecialty program where you can offer your skills, rather than the other way around, may be the more sustainable, long-term option.” TH

Gretchen Henkel is a frequent contributor to The Hospitalist.

References

1. Ogershok PR, Li X, Palmer HC, et al. Restructuring an academic pediatric inpatient service using concepts developed by hospitalists.” Clin Pediatr. 2001 Dec.;40(12): 653-660.
2. Tenner PA, Dibrell H, Taylor RP. Improved survival with hospitalists in a pediatric intensive care unit.” Crit Care Med. 2003 Mar;31(3):847-852.
3. Melzer SM, Molteni, RA, Marcuse EK, et al. Characteristics and financial performance of a pediatric faculty inpatient attending service: a resource-based relative value scale analysis. Pediatrics. 2001 Jul;108(1);79-84.

Woman claims improper contact during treatment

Fairfax County (VA) Circuit Court

A 23-year-old woman who received treatment from a psychiatrist for approximately 2½ years claimed that he sexually abused her during that time. She alleged that the inappropriate sexual relationship included holding, hugging, kissing, fondling, and watching pornography. The patient claimed that the relationship led to emotional distress and caused her to attempt suicide.

The psychiatrist admitted that a sexual relationship occurred but contended that the patient suffered no harm.

A $400,000 verdict was returned

Did inappropriate contact cause agoraphobia, anorexia?

Suffolk County (MA) Superior Court

A patient in her 20s had a history of emotional problems and sexual assaults against her. A psychiatrist treated her for obsessive-compulsive disorder for 4 years. He acknowledged giving the patient stuffed animals, cards, and letters and visiting her home several times when she was unable to go to his office. During sessions he touched her hand for comfort and hugged her. The patient claimed they had regular sexual contact.

The patient alleged that the psychiatrist was negligent for engaging in inappropriate sexual conduct, which she claims caused ongoing emotional distress. She claimed she was unable to work and suffered from agoraphobia, intimate relationships difficulties, and anorexia as a result of his actions. The psychiatrist denied any inappropriate sexual conduct.

The psychiatrist’s license was suspended indefinitely, but the suspension was stayed under an agreement that he attend medical education courses.

A $750,000 settlement was reached

Dr. Grant’s observations

Although most physicians would agree that sexual relations with a patient are inappropriate,¹ the fact that cases continue to occur suggests a need to emphasize treatment boundaries. Establishing clear boundaries in the doctor-patient relationship creates an atmosphere of safety and predictability that allows treatment to thrive.²

Boundary problems are one of the most frequent reasons for malpractice actions against mental health providers.³ Although much of the literature discusses boundary violations during psychotherapy, issues may arise in all treatment settings, including psychopharmacologic management.

One-half of all psychiatrists will treat at least 1 victim of physician sexual misconduct during their careers.⁴ One study⁵ examining sex-related offenses committed by U.S. physicians in all specialties found:

• The number of physicians disciplined for sex-related offenses increased each year from 1989 to 1996.
• 22% of disciplined physicians had sexual intercourse with patients, 15% had sexual contact or touching, 37% committed other sexual abuse that did not fit in either of these 2 categories, and 25% involved nonpatients.
• 28% of disciplined physicians were psychiatrists, the most represented specialty in the study.

Box

In a 1986 survey of psychiatrists, 7% of male and 3% of female clinicians reported having sexual contact with their patients.⁶ A 1988 survey of fourth-year psychiatry residents found that 1% of respondents acknowledged having sexual relations with a patient.⁷ In a 1992 study, 9% of physicians across specialties reported engaging in sexual contact with 1 or more current or former patients.⁸ In that study, 19% of female physicians and 40% of male physicians reported that they did not think physician-patient sexual misconduct was always harmful to patients.⁸ These views and behaviors are in violation of medical codes of ethics (Box).⁹

How misconduct harms patients

Trust is essential to establishing a secure therapeutic relationship. Boundary violations may result in missed diagnoses, inappropriate treatment, and/or worsened psychiatric symptoms. Patients might develop complex posttraumatic stress disorder, depression, anxiety, dissociation, sexual dysfunction, somatoform disorders, eating disorders, sleep disorders, or substance use disorders.⁴ They could lose faith in their treatment providers, have difficulties expressing anger, feel guilty, develop poor self-concept, experience a loss of confidence, and develop problems establishing trusting relationships.⁴ For these reasons, clinicians can be sued for negligent treatment and sexual misconduct.¹⁰

Boundary violations

Although sexual activities with patients are clear boundary violations, what about the second case when the therapist gave the patient stuffed animals and cards and hugged her? Progressive boundary violations often precede and accompany sexual misconduct.¹⁰

Five risk factors have been associated with therapist boundary violations:³

• life crises—effects of aging, career disappointments, unfulfilled hopes, or marital conflicts
• transitions—job changes or job loss
• medical illness
• arrogance—the belief that a boundary violation couldn’t happen to him or her and not recognizing the need for consultation
• common stress points with the patient

Although the list is not exhaustive, these factors may be associated with a psychiatrist turning to the patient for solace, gratification, or excitement.

Drawing boundary lines

Not all boundary issues are the same, and Gutheil et al² suggest 2 categories:

• Boundary crossings—a benign variant where the deviation may advance therapy in a constructive way that does not harm the patient, such as discussion of countertransference.
• Boundary violations—the transgression harms or exploits the patient.

Although some boundary issues may appear benign, even theoretically harmless boundary crossings can be misrepresented or misconstrued by the patient.¹¹ Also, boundary transgressions that do not involve erotic touch might harm the treatment process and the patient.²

When examining “minor” boundary issues that may seem innocuous, ask yourself if the action is for your benefit rather than to advance the patient’s therapy. Also, determine if the intervention is part of a series of progressive boundary violations. If the answer to either question is “yes,” desist immediately and take corrective action.¹⁰

The psychiatrist has a professional code of ethics to follow and can be held responsible for failing to set or adhere to boundaries.¹¹ If a patient initiates a boundary violation, you must refuse and then explore the patient’s underlying psychological issues, perhaps aided by consultation with a peer or mentor (Table). Repeated patient demands to breach boundaries requires prompt consultation to determine if you can continue treating the patient or if you should transfer the patient to another clinician. Document the patient’s demands to breach boundaries and your actions when seeking consultation.³

Table 1

How to maintain integrity of the treatment process

Woman claims improper contact during treatment

Fairfax County (VA) Circuit Court

A 23-year-old woman who received treatment from a psychiatrist for approximately 2½ years claimed that he sexually abused her during that time. She alleged that the inappropriate sexual relationship included holding, hugging, kissing, fondling, and watching pornography. The patient claimed that the relationship led to emotional distress and caused her to attempt suicide.

The psychiatrist admitted that a sexual relationship occurred but contended that the patient suffered no harm.

A $400,000 verdict was returned

Did inappropriate contact cause agoraphobia, anorexia?

Suffolk County (MA) Superior Court

A patient in her 20s had a history of emotional problems and sexual assaults against her. A psychiatrist treated her for obsessive-compulsive disorder for 4 years. He acknowledged giving the patient stuffed animals, cards, and letters and visiting her home several times when she was unable to go to his office. During sessions he touched her hand for comfort and hugged her. The patient claimed they had regular sexual contact.

The patient alleged that the psychiatrist was negligent for engaging in inappropriate sexual conduct, which she claims caused ongoing emotional distress. She claimed she was unable to work and suffered from agoraphobia, intimate relationships difficulties, and anorexia as a result of his actions. The psychiatrist denied any inappropriate sexual conduct.

The psychiatrist’s license was suspended indefinitely, but the suspension was stayed under an agreement that he attend medical education courses.

A $750,000 settlement was reached

Dr. Grant’s observations

Although most physicians would agree that sexual relations with a patient are inappropriate,¹ the fact that cases continue to occur suggests a need to emphasize treatment boundaries. Establishing clear boundaries in the doctor-patient relationship creates an atmosphere of safety and predictability that allows treatment to thrive.²

Boundary problems are one of the most frequent reasons for malpractice actions against mental health providers.³ Although much of the literature discusses boundary violations during psychotherapy, issues may arise in all treatment settings, including psychopharmacologic management.

One-half of all psychiatrists will treat at least 1 victim of physician sexual misconduct during their careers.⁴ One study⁵ examining sex-related offenses committed by U.S. physicians in all specialties found:

• The number of physicians disciplined for sex-related offenses increased each year from 1989 to 1996.
• 22% of disciplined physicians had sexual intercourse with patients, 15% had sexual contact or touching, 37% committed other sexual abuse that did not fit in either of these 2 categories, and 25% involved nonpatients.
• 28% of disciplined physicians were psychiatrists, the most represented specialty in the study.

Box

In a 1986 survey of psychiatrists, 7% of male and 3% of female clinicians reported having sexual contact with their patients.⁶ A 1988 survey of fourth-year psychiatry residents found that 1% of respondents acknowledged having sexual relations with a patient.⁷ In a 1992 study, 9% of physicians across specialties reported engaging in sexual contact with 1 or more current or former patients.⁸ In that study, 19% of female physicians and 40% of male physicians reported that they did not think physician-patient sexual misconduct was always harmful to patients.⁸ These views and behaviors are in violation of medical codes of ethics (Box).⁹

How misconduct harms patients

Trust is essential to establishing a secure therapeutic relationship. Boundary violations may result in missed diagnoses, inappropriate treatment, and/or worsened psychiatric symptoms. Patients might develop complex posttraumatic stress disorder, depression, anxiety, dissociation, sexual dysfunction, somatoform disorders, eating disorders, sleep disorders, or substance use disorders.⁴ They could lose faith in their treatment providers, have difficulties expressing anger, feel guilty, develop poor self-concept, experience a loss of confidence, and develop problems establishing trusting relationships.⁴ For these reasons, clinicians can be sued for negligent treatment and sexual misconduct.¹⁰

Boundary violations

Although sexual activities with patients are clear boundary violations, what about the second case when the therapist gave the patient stuffed animals and cards and hugged her? Progressive boundary violations often precede and accompany sexual misconduct.¹⁰

Five risk factors have been associated with therapist boundary violations:³

• life crises—effects of aging, career disappointments, unfulfilled hopes, or marital conflicts
• transitions—job changes or job loss
• medical illness
• arrogance—the belief that a boundary violation couldn’t happen to him or her and not recognizing the need for consultation
• common stress points with the patient

Although the list is not exhaustive, these factors may be associated with a psychiatrist turning to the patient for solace, gratification, or excitement.

Drawing boundary lines

Not all boundary issues are the same, and Gutheil et al² suggest 2 categories:

• Boundary crossings—a benign variant where the deviation may advance therapy in a constructive way that does not harm the patient, such as discussion of countertransference.
• Boundary violations—the transgression harms or exploits the patient.

Although some boundary issues may appear benign, even theoretically harmless boundary crossings can be misrepresented or misconstrued by the patient.¹¹ Also, boundary transgressions that do not involve erotic touch might harm the treatment process and the patient.²

When examining “minor” boundary issues that may seem innocuous, ask yourself if the action is for your benefit rather than to advance the patient’s therapy. Also, determine if the intervention is part of a series of progressive boundary violations. If the answer to either question is “yes,” desist immediately and take corrective action.¹⁰

The psychiatrist has a professional code of ethics to follow and can be held responsible for failing to set or adhere to boundaries.¹¹ If a patient initiates a boundary violation, you must refuse and then explore the patient’s underlying psychological issues, perhaps aided by consultation with a peer or mentor (Table). Repeated patient demands to breach boundaries requires prompt consultation to determine if you can continue treating the patient or if you should transfer the patient to another clinician. Document the patient’s demands to breach boundaries and your actions when seeking consultation.³

Table 1

How to maintain integrity of the treatment process

Woman claims improper contact during treatment

Fairfax County (VA) Circuit Court

A 23-year-old woman who received treatment from a psychiatrist for approximately 2½ years claimed that he sexually abused her during that time. She alleged that the inappropriate sexual relationship included holding, hugging, kissing, fondling, and watching pornography. The patient claimed that the relationship led to emotional distress and caused her to attempt suicide.

The psychiatrist admitted that a sexual relationship occurred but contended that the patient suffered no harm.

A $400,000 verdict was returned

Did inappropriate contact cause agoraphobia, anorexia?

Suffolk County (MA) Superior Court

A patient in her 20s had a history of emotional problems and sexual assaults against her. A psychiatrist treated her for obsessive-compulsive disorder for 4 years. He acknowledged giving the patient stuffed animals, cards, and letters and visiting her home several times when she was unable to go to his office. During sessions he touched her hand for comfort and hugged her. The patient claimed they had regular sexual contact.

The patient alleged that the psychiatrist was negligent for engaging in inappropriate sexual conduct, which she claims caused ongoing emotional distress. She claimed she was unable to work and suffered from agoraphobia, intimate relationships difficulties, and anorexia as a result of his actions. The psychiatrist denied any inappropriate sexual conduct.

The psychiatrist’s license was suspended indefinitely, but the suspension was stayed under an agreement that he attend medical education courses.

A $750,000 settlement was reached

Dr. Grant’s observations

Although most physicians would agree that sexual relations with a patient are inappropriate,¹ the fact that cases continue to occur suggests a need to emphasize treatment boundaries. Establishing clear boundaries in the doctor-patient relationship creates an atmosphere of safety and predictability that allows treatment to thrive.²

Boundary problems are one of the most frequent reasons for malpractice actions against mental health providers.³ Although much of the literature discusses boundary violations during psychotherapy, issues may arise in all treatment settings, including psychopharmacologic management.

One-half of all psychiatrists will treat at least 1 victim of physician sexual misconduct during their careers.⁴ One study⁵ examining sex-related offenses committed by U.S. physicians in all specialties found:

• The number of physicians disciplined for sex-related offenses increased each year from 1989 to 1996.
• 22% of disciplined physicians had sexual intercourse with patients, 15% had sexual contact or touching, 37% committed other sexual abuse that did not fit in either of these 2 categories, and 25% involved nonpatients.
• 28% of disciplined physicians were psychiatrists, the most represented specialty in the study.

Box

In a 1986 survey of psychiatrists, 7% of male and 3% of female clinicians reported having sexual contact with their patients.⁶ A 1988 survey of fourth-year psychiatry residents found that 1% of respondents acknowledged having sexual relations with a patient.⁷ In a 1992 study, 9% of physicians across specialties reported engaging in sexual contact with 1 or more current or former patients.⁸ In that study, 19% of female physicians and 40% of male physicians reported that they did not think physician-patient sexual misconduct was always harmful to patients.⁸ These views and behaviors are in violation of medical codes of ethics (Box).⁹

How misconduct harms patients

Trust is essential to establishing a secure therapeutic relationship. Boundary violations may result in missed diagnoses, inappropriate treatment, and/or worsened psychiatric symptoms. Patients might develop complex posttraumatic stress disorder, depression, anxiety, dissociation, sexual dysfunction, somatoform disorders, eating disorders, sleep disorders, or substance use disorders.⁴ They could lose faith in their treatment providers, have difficulties expressing anger, feel guilty, develop poor self-concept, experience a loss of confidence, and develop problems establishing trusting relationships.⁴ For these reasons, clinicians can be sued for negligent treatment and sexual misconduct.¹⁰

Boundary violations

Although sexual activities with patients are clear boundary violations, what about the second case when the therapist gave the patient stuffed animals and cards and hugged her? Progressive boundary violations often precede and accompany sexual misconduct.¹⁰

Five risk factors have been associated with therapist boundary violations:³

• life crises—effects of aging, career disappointments, unfulfilled hopes, or marital conflicts
• transitions—job changes or job loss
• medical illness
• arrogance—the belief that a boundary violation couldn’t happen to him or her and not recognizing the need for consultation
• common stress points with the patient

Although the list is not exhaustive, these factors may be associated with a psychiatrist turning to the patient for solace, gratification, or excitement.

Drawing boundary lines

Not all boundary issues are the same, and Gutheil et al² suggest 2 categories:

• Boundary crossings—a benign variant where the deviation may advance therapy in a constructive way that does not harm the patient, such as discussion of countertransference.
• Boundary violations—the transgression harms or exploits the patient.

Although some boundary issues may appear benign, even theoretically harmless boundary crossings can be misrepresented or misconstrued by the patient.¹¹ Also, boundary transgressions that do not involve erotic touch might harm the treatment process and the patient.²

When examining “minor” boundary issues that may seem innocuous, ask yourself if the action is for your benefit rather than to advance the patient’s therapy. Also, determine if the intervention is part of a series of progressive boundary violations. If the answer to either question is “yes,” desist immediately and take corrective action.¹⁰

The psychiatrist has a professional code of ethics to follow and can be held responsible for failing to set or adhere to boundaries.¹¹ If a patient initiates a boundary violation, you must refuse and then explore the patient’s underlying psychological issues, perhaps aided by consultation with a peer or mentor (Table). Repeated patient demands to breach boundaries requires prompt consultation to determine if you can continue treating the patient or if you should transfer the patient to another clinician. Document the patient’s demands to breach boundaries and your actions when seeking consultation.³

Table 1

How to maintain integrity of the treatment process

Q We’re seeing more and more patients who are obese. Because of their habitus, we are unable to evaluate the uterus and ovaries adequately at the time of the well-woman or diagnostic examination. In such cases, we’ve begun ordering transvaginal ultrasonography (US). Our billing staff reports that most of these claims are being denied for lack of medical necessity. Any suggestions?

A This is a dilemma, to be sure—one where the payer has the deciding voice over what is and what isn’t medically indicated. An option is to report a diagnosis of obesity linked to the US scan, along with V72.31 for a preventive exam or another diagnosis code representing the presenting problem or complaint.

If the US scan is also inconclusive, you can report 793.91 (image test inconclusive due to excess body fat) with another code that represents the patient’s documented body mass index.

Appeal the determination if the US scan is denied on the basis of medical necessity. Let the payer know that a thorough pelvic exam could not be completed because of the patient’s body mass; point out that obesity is a risk factor for cancer.

Making those points should help you get paid, eventually—although The Centers for Medicare & Medicaid Services (CMS) and most payers have determined that US as a confirmatory adjunct to physical examination (in the absence of an abnormal finding) will not be reimbursed.

Slow payment for unlisted codes for lap hysterectomy

Q My surgeon performed a total laparoscopic hysterectomy in which he removed the entire specimen through the laparoscope. Must I report an unlisted code for this procedure?

A No. You have the option instead of reporting one of the codes for a laparoscopic vaginal hysterectomy (codes 58550–58554) because the vaginal part is only for retrieving the specimen, which is otherwise released from its attachment through the laparoscope. This is the recommendation of ACOG and the American Association of Gynecologic Laparoscopists (AAGL).

(There is good news here: New codes for total laparoscopic hysterectomy will be available beginning January 1, 2008. Look for details on these and other changes in Reimbursement Adviser in the December 2007 issue of OBG Management.)

Of course, your other coding option is to report the unlisted code 58578 (un-listed laparoscopic procedure, uterus). If you choose that option, however, you will have to submit the operative report with the claim, along with a letter from the physician explaining why this procedure was more advantageous to the patient than LAVH. And although using an unlisted code will not mean denial of service, it will slow down payment.

No need for modifiers on self-performed US scans

Q Our four-physician OB practice performs limited US scans on our pregnant patients. The only code we use is 76815. The practice owns the US machine and all four of us perform the scans, print the photographs, and create reports from the machine ourselves. Should we be billing these scans with modifier -26?

A No. Modifier -26 is only reported when the global service is not provided—that is, when you do not own the US machine. The unmodified code 76815 represents the technical and professional components of the US procedure, so you are coding correctly by not adding a modifier.

Although you are acting as your own sonographer for your scans, this still represents the technical component of a scan, which is reimbursed when reporting the unmodified code for the service.

It’s “false labor” if there’s no bleeding—at any date

Q My pregnant patient who delivered her previous pregnancy at 28 weeks because of premature labor is now complaining of contractions at 20.3 weeks. Would 640.03 (threatened abortion; antepartum condition or complication) be the appropriate code even though she is not bleeding?

A Twenty weeks is very early to deliver, but you would have to report 644.13 (other threatened labor) because it is the default code for false labor regardless of gestational weeks, according to a staff member of the ICD-9-CM Coordination and Maintenance Committee. A hemorrhage code, such as 640.03, should not be reported in the absence of documented bleeding. If contractions progress, however, move on to the next code that matches the situation.

Colporrhaphy? Do not code for posterior repair

Q We have been told that we can report code 45560 (repair of rectocele [separate procedure]) for posterior repair of a rectocele. I’ve noted that the relative value units (RVUs) for this procedure are higher than for a posterior colporrhaphy. Please clarify: When is it appropriate to bill 45560?

A The simple answer is that you must bill the procedure that you’ve documented, and colporrhaphy is the procedure performed by 99% of ObGyns to repair a rectocele. Typically, this involves making a midline incision in the posterior vaginal wall, plicating rectovaginal tissue, suturing it together, cutting off excess tissue, and, sometimes, supporting weakened rectovaginal tissue with mesh.

The code 45560, on the other hand, is listed in the digestive section of CPT and is, basically, a transanal approach procedure that has a vaginal component. It is much different than posterior colporrhaphy, and is typically performed when a patient with a rectocele has fecal incontinence.

Although 45560 does carry slightly more RVUs than 57250 (posterior colporrhaphy, repair of rectocele with or without perineorrhaphy), the physician work portion of the RVU total for both codes is now identical: 11.42.

Last, be aware: Payers consider it fraud for a surgeon to use the wrong code in this situation because it pays more, when, in fact, colporrhaphy has been performed and documented.

Patient asks for test; is that “medical necessity”?

Q Occasionally, we see a patient with a family history of ovarian cancer who requests a test for cancer antigen 125. If the result is elevated and we decide to perform a US scan, what diagnosis code should we add for medical necessity? Our experience using a family history code with payers has not been good.

A The correct primary diagnosis code for this situation is 795.82 (elevated cancer antigen 125), followed by V16.41 (family history of malignant neoplasm; genital organs; ovary).

Q We’re seeing more and more patients who are obese. Because of their habitus, we are unable to evaluate the uterus and ovaries adequately at the time of the well-woman or diagnostic examination. In such cases, we’ve begun ordering transvaginal ultrasonography (US). Our billing staff reports that most of these claims are being denied for lack of medical necessity. Any suggestions?

A This is a dilemma, to be sure—one where the payer has the deciding voice over what is and what isn’t medically indicated. An option is to report a diagnosis of obesity linked to the US scan, along with V72.31 for a preventive exam or another diagnosis code representing the presenting problem or complaint.

If the US scan is also inconclusive, you can report 793.91 (image test inconclusive due to excess body fat) with another code that represents the patient’s documented body mass index.

Appeal the determination if the US scan is denied on the basis of medical necessity. Let the payer know that a thorough pelvic exam could not be completed because of the patient’s body mass; point out that obesity is a risk factor for cancer.

Making those points should help you get paid, eventually—although The Centers for Medicare & Medicaid Services (CMS) and most payers have determined that US as a confirmatory adjunct to physical examination (in the absence of an abnormal finding) will not be reimbursed.

Slow payment for unlisted codes for lap hysterectomy

Q My surgeon performed a total laparoscopic hysterectomy in which he removed the entire specimen through the laparoscope. Must I report an unlisted code for this procedure?

A No. You have the option instead of reporting one of the codes for a laparoscopic vaginal hysterectomy (codes 58550–58554) because the vaginal part is only for retrieving the specimen, which is otherwise released from its attachment through the laparoscope. This is the recommendation of ACOG and the American Association of Gynecologic Laparoscopists (AAGL).

(There is good news here: New codes for total laparoscopic hysterectomy will be available beginning January 1, 2008. Look for details on these and other changes in Reimbursement Adviser in the December 2007 issue of OBG Management.)

Of course, your other coding option is to report the unlisted code 58578 (un-listed laparoscopic procedure, uterus). If you choose that option, however, you will have to submit the operative report with the claim, along with a letter from the physician explaining why this procedure was more advantageous to the patient than LAVH. And although using an unlisted code will not mean denial of service, it will slow down payment.

No need for modifiers on self-performed US scans

Q Our four-physician OB practice performs limited US scans on our pregnant patients. The only code we use is 76815. The practice owns the US machine and all four of us perform the scans, print the photographs, and create reports from the machine ourselves. Should we be billing these scans with modifier -26?

A No. Modifier -26 is only reported when the global service is not provided—that is, when you do not own the US machine. The unmodified code 76815 represents the technical and professional components of the US procedure, so you are coding correctly by not adding a modifier.

Although you are acting as your own sonographer for your scans, this still represents the technical component of a scan, which is reimbursed when reporting the unmodified code for the service.

It’s “false labor” if there’s no bleeding—at any date

Q My pregnant patient who delivered her previous pregnancy at 28 weeks because of premature labor is now complaining of contractions at 20.3 weeks. Would 640.03 (threatened abortion; antepartum condition or complication) be the appropriate code even though she is not bleeding?

A Twenty weeks is very early to deliver, but you would have to report 644.13 (other threatened labor) because it is the default code for false labor regardless of gestational weeks, according to a staff member of the ICD-9-CM Coordination and Maintenance Committee. A hemorrhage code, such as 640.03, should not be reported in the absence of documented bleeding. If contractions progress, however, move on to the next code that matches the situation.

Colporrhaphy? Do not code for posterior repair

Q We have been told that we can report code 45560 (repair of rectocele [separate procedure]) for posterior repair of a rectocele. I’ve noted that the relative value units (RVUs) for this procedure are higher than for a posterior colporrhaphy. Please clarify: When is it appropriate to bill 45560?

A The simple answer is that you must bill the procedure that you’ve documented, and colporrhaphy is the procedure performed by 99% of ObGyns to repair a rectocele. Typically, this involves making a midline incision in the posterior vaginal wall, plicating rectovaginal tissue, suturing it together, cutting off excess tissue, and, sometimes, supporting weakened rectovaginal tissue with mesh.

The code 45560, on the other hand, is listed in the digestive section of CPT and is, basically, a transanal approach procedure that has a vaginal component. It is much different than posterior colporrhaphy, and is typically performed when a patient with a rectocele has fecal incontinence.

Although 45560 does carry slightly more RVUs than 57250 (posterior colporrhaphy, repair of rectocele with or without perineorrhaphy), the physician work portion of the RVU total for both codes is now identical: 11.42.

Last, be aware: Payers consider it fraud for a surgeon to use the wrong code in this situation because it pays more, when, in fact, colporrhaphy has been performed and documented.

Patient asks for test; is that “medical necessity”?

Q Occasionally, we see a patient with a family history of ovarian cancer who requests a test for cancer antigen 125. If the result is elevated and we decide to perform a US scan, what diagnosis code should we add for medical necessity? Our experience using a family history code with payers has not been good.

A The correct primary diagnosis code for this situation is 795.82 (elevated cancer antigen 125), followed by V16.41 (family history of malignant neoplasm; genital organs; ovary).

Q We’re seeing more and more patients who are obese. Because of their habitus, we are unable to evaluate the uterus and ovaries adequately at the time of the well-woman or diagnostic examination. In such cases, we’ve begun ordering transvaginal ultrasonography (US). Our billing staff reports that most of these claims are being denied for lack of medical necessity. Any suggestions?

A This is a dilemma, to be sure—one where the payer has the deciding voice over what is and what isn’t medically indicated. An option is to report a diagnosis of obesity linked to the US scan, along with V72.31 for a preventive exam or another diagnosis code representing the presenting problem or complaint.

If the US scan is also inconclusive, you can report 793.91 (image test inconclusive due to excess body fat) with another code that represents the patient’s documented body mass index.

Appeal the determination if the US scan is denied on the basis of medical necessity. Let the payer know that a thorough pelvic exam could not be completed because of the patient’s body mass; point out that obesity is a risk factor for cancer.

Making those points should help you get paid, eventually—although The Centers for Medicare & Medicaid Services (CMS) and most payers have determined that US as a confirmatory adjunct to physical examination (in the absence of an abnormal finding) will not be reimbursed.

Slow payment for unlisted codes for lap hysterectomy

Q My surgeon performed a total laparoscopic hysterectomy in which he removed the entire specimen through the laparoscope. Must I report an unlisted code for this procedure?

A No. You have the option instead of reporting one of the codes for a laparoscopic vaginal hysterectomy (codes 58550–58554) because the vaginal part is only for retrieving the specimen, which is otherwise released from its attachment through the laparoscope. This is the recommendation of ACOG and the American Association of Gynecologic Laparoscopists (AAGL).

(There is good news here: New codes for total laparoscopic hysterectomy will be available beginning January 1, 2008. Look for details on these and other changes in Reimbursement Adviser in the December 2007 issue of OBG Management.)

Of course, your other coding option is to report the unlisted code 58578 (un-listed laparoscopic procedure, uterus). If you choose that option, however, you will have to submit the operative report with the claim, along with a letter from the physician explaining why this procedure was more advantageous to the patient than LAVH. And although using an unlisted code will not mean denial of service, it will slow down payment.

No need for modifiers on self-performed US scans

Q Our four-physician OB practice performs limited US scans on our pregnant patients. The only code we use is 76815. The practice owns the US machine and all four of us perform the scans, print the photographs, and create reports from the machine ourselves. Should we be billing these scans with modifier -26?

A No. Modifier -26 is only reported when the global service is not provided—that is, when you do not own the US machine. The unmodified code 76815 represents the technical and professional components of the US procedure, so you are coding correctly by not adding a modifier.

Although you are acting as your own sonographer for your scans, this still represents the technical component of a scan, which is reimbursed when reporting the unmodified code for the service.

It’s “false labor” if there’s no bleeding—at any date

Q My pregnant patient who delivered her previous pregnancy at 28 weeks because of premature labor is now complaining of contractions at 20.3 weeks. Would 640.03 (threatened abortion; antepartum condition or complication) be the appropriate code even though she is not bleeding?

A Twenty weeks is very early to deliver, but you would have to report 644.13 (other threatened labor) because it is the default code for false labor regardless of gestational weeks, according to a staff member of the ICD-9-CM Coordination and Maintenance Committee. A hemorrhage code, such as 640.03, should not be reported in the absence of documented bleeding. If contractions progress, however, move on to the next code that matches the situation.

Colporrhaphy? Do not code for posterior repair

Q We have been told that we can report code 45560 (repair of rectocele [separate procedure]) for posterior repair of a rectocele. I’ve noted that the relative value units (RVUs) for this procedure are higher than for a posterior colporrhaphy. Please clarify: When is it appropriate to bill 45560?

A The simple answer is that you must bill the procedure that you’ve documented, and colporrhaphy is the procedure performed by 99% of ObGyns to repair a rectocele. Typically, this involves making a midline incision in the posterior vaginal wall, plicating rectovaginal tissue, suturing it together, cutting off excess tissue, and, sometimes, supporting weakened rectovaginal tissue with mesh.

The code 45560, on the other hand, is listed in the digestive section of CPT and is, basically, a transanal approach procedure that has a vaginal component. It is much different than posterior colporrhaphy, and is typically performed when a patient with a rectocele has fecal incontinence.

Although 45560 does carry slightly more RVUs than 57250 (posterior colporrhaphy, repair of rectocele with or without perineorrhaphy), the physician work portion of the RVU total for both codes is now identical: 11.42.

Last, be aware: Payers consider it fraud for a surgeon to use the wrong code in this situation because it pays more, when, in fact, colporrhaphy has been performed and documented.

Patient asks for test; is that “medical necessity”?

Q Occasionally, we see a patient with a family history of ovarian cancer who requests a test for cancer antigen 125. If the result is elevated and we decide to perform a US scan, what diagnosis code should we add for medical necessity? Our experience using a family history code with payers has not been good.

A The correct primary diagnosis code for this situation is 795.82 (elevated cancer antigen 125), followed by V16.41 (family history of malignant neoplasm; genital organs; ovary).

Evidence summary

A solitary pulmonary nodule, or “coin lesion,” is an intraparenchymal finding on chest radiograph or CT that is less than 3 to 4 cm in diameter and not associated with atelectasis or adenopathy. Malignancy rates range from 15% to 75%, depending on the population studied.¹ Although early detection of malignancy portends a major improvement in survival (up to 75% at 5 years following surgical resection of stage IA disease), most lung cancers progress asymptomatically until quite advanced.²

The presumed benign nature of lesions that are either unchanged over 2 years or have central calcifications is based on 3 retrospective studies from the 1950s.^3-6 However, these should not be considered absolutes. A recent study revisiting the original data calculated the predictive value of benign nature based on no growth to be only 65% (95% confidence interval [CI], 47%–83%).⁷ Also, a study assessing the accuracy of radiologists’ assessment of calcification in solitary pulmonary nodules compared with thin-section CT found that 7% of “definitely calcified” nodules on chest radiograph lacked calcification on thin-section CT.⁸

Which clinical variables best predict malignancy?

The best available clinical decision rule was derived and validated from a single split population of patients with solitary pulmonary nodules.⁹ The outcome variable was defined as malignancy based on histologic tissue analysis or benignity by radiographic stability or resolution over 2 years. The authors did not report whether those determining outcomes and predictors were appropriately blinded.

The authors found that 3 clinical variables (age, smoking history, and cancer history) plus 3 radiographic variables (diameter, spiculation, and nodule location in the upper lobes) were independent predictors of malignancy. An online calculator using this prediction model is available at www.chestx-ray.com/SPN/ SPNProb.html.¹⁰

CT or PET?

Three comparative studies observed 8 to 12 radiologists’ readings of high-resolution CT images of 28 to 56 patients with solitary pulmonary nodules (established diagnoses by either histology or stability over time).^11-13 Approximately half the nodules represented malignant lesions.

Radiologists assigned a level of confidence to their assessment of each case as benign or malignant. At a minimum, they were informed of each patient’s age and gender, and in 2 studies they also knew other information, such as the patient’s smoking and cancer histories. The study showed that the radiologists would have correctly diagnosed a pair of solitary pulmonary nodule cases, one malignant and one benign, between 75% and 83% of the time. Conversely, 17% to 25% of the time they would have diagnosed the case pair incorrectly.

A meta-analysis of 40 studies of FDG PET scanning for solitary pulmonary nodules yielded a maximum joint sensitivity and specificity of 90% (95% CI, 86.4%– 92.7%).¹⁴ The methodological quality of studies included in the meta-analysis was fair, with small sample sizes (inclusion criteria were for a minimum of 10 patients with pulmonary nodules and malignant prevalence of at least 0.5); masking was frequently incomplete.

Sensitivity of histologic/cytologic tests varies

A recent systematic review of studies evaluating patients with suspected lung cancer looked into the diagnostic sensitivity of various methods of histologic and cytologic tests.¹⁵ Researchers compared the evaluated test results to a reference standard of pathology/histology, definitive cytology, or at least 1-year radiographic follow-up.

Transbronchial needle aspiration showed a sensitivity of 67% (95% CI, 64%–70%) for peripheral lung malignancy of any size; however, only 5 studies met study criteria and their sample sizes varied greatly (n=20 to n=480). Eight studies looking at bronchoscopy (including brush or biopsy) for peripheral lung lesions <2 cm in diameter yielded a sensitivity of only 33% (95% CI, 28%–38%). In the same systematic review, 61 studies of transthoracic needle aspiration for localized pulmonary lesions of any size had a pooled sensitivity of 90% (95% CI, 88%–92%). The prevalence of malignancy in the studies ranged from 0.58 to 0.93.¹⁵ Factors affecting heterogeneity between studies included the wide range in study dates, imaging technology used, and study sizes.

What test is most cost-effective?

CT appears cost-effective when the pretest probability of malignancy is <90%; therefore, consider it on virtually all new cases of solitary pulmonary nodules.¹ Also, when CT and pretest risk-assessments are discordant (eg, a patient has a low pretest probability of malignancy but his CT is suggestive of malignancy), the FDG PET scan is the most economically feasible at less than $20,000 per quality-adjusted life year.

Recommendations from others

The American College of Chest Physicians (ACCP)² suggests pursuing no further evaluation if a nodule is unchanged for >2 years or has benign central calcifications. They recommend that physicians perform CT on every patient with a new nodule to characterize the nodule, its location, and the mediastinum. They do not recommend PET scans for nodules <1 cm. Patients who are marginal surgical candidates and have a negative PET scan should have a repeat CT scan in 3 months; serial CTs at 3, 6, 12, and 24 months are suggested, too, if prior chest radiographs are negative.

The ACCP states that transthoracic needle aspiration is not indicated in surgical candidates unless they decline surgery; then transthoracic needle aspiration or a transbronchial approach are the preferred procedure. Transthoracic needle aspiration may also be useful in establishing a diagnosis for patients who are not surgical candidates or who have a high surgical risk.

ACCP expert consensus favors the reference standard of video-assisted thoracoscopic surgery with wedge resection as the ideal method for obtaining tissue diagnosis in consenting, operable patients with solitary pulmonary nodules. Objective evidence is lacking on follow-up monitoring methods for patients with a nodule who do not have a tissue diagnosis and observation alone is chosen. ACCP expert consensus favors a 2-year follow-up with CT scanning at 3, 6, 12, and 24 months to monitor for nodule growth.²

Evidence summary

A solitary pulmonary nodule, or “coin lesion,” is an intraparenchymal finding on chest radiograph or CT that is less than 3 to 4 cm in diameter and not associated with atelectasis or adenopathy. Malignancy rates range from 15% to 75%, depending on the population studied.¹ Although early detection of malignancy portends a major improvement in survival (up to 75% at 5 years following surgical resection of stage IA disease), most lung cancers progress asymptomatically until quite advanced.²

The presumed benign nature of lesions that are either unchanged over 2 years or have central calcifications is based on 3 retrospective studies from the 1950s.^3-6 However, these should not be considered absolutes. A recent study revisiting the original data calculated the predictive value of benign nature based on no growth to be only 65% (95% confidence interval [CI], 47%–83%).⁷ Also, a study assessing the accuracy of radiologists’ assessment of calcification in solitary pulmonary nodules compared with thin-section CT found that 7% of “definitely calcified” nodules on chest radiograph lacked calcification on thin-section CT.⁸

Which clinical variables best predict malignancy?

The best available clinical decision rule was derived and validated from a single split population of patients with solitary pulmonary nodules.⁹ The outcome variable was defined as malignancy based on histologic tissue analysis or benignity by radiographic stability or resolution over 2 years. The authors did not report whether those determining outcomes and predictors were appropriately blinded.

The authors found that 3 clinical variables (age, smoking history, and cancer history) plus 3 radiographic variables (diameter, spiculation, and nodule location in the upper lobes) were independent predictors of malignancy. An online calculator using this prediction model is available at www.chestx-ray.com/SPN/ SPNProb.html.¹⁰

CT or PET?

Three comparative studies observed 8 to 12 radiologists’ readings of high-resolution CT images of 28 to 56 patients with solitary pulmonary nodules (established diagnoses by either histology or stability over time).^11-13 Approximately half the nodules represented malignant lesions.

Radiologists assigned a level of confidence to their assessment of each case as benign or malignant. At a minimum, they were informed of each patient’s age and gender, and in 2 studies they also knew other information, such as the patient’s smoking and cancer histories. The study showed that the radiologists would have correctly diagnosed a pair of solitary pulmonary nodule cases, one malignant and one benign, between 75% and 83% of the time. Conversely, 17% to 25% of the time they would have diagnosed the case pair incorrectly.

A meta-analysis of 40 studies of FDG PET scanning for solitary pulmonary nodules yielded a maximum joint sensitivity and specificity of 90% (95% CI, 86.4%– 92.7%).¹⁴ The methodological quality of studies included in the meta-analysis was fair, with small sample sizes (inclusion criteria were for a minimum of 10 patients with pulmonary nodules and malignant prevalence of at least 0.5); masking was frequently incomplete.

Sensitivity of histologic/cytologic tests varies

A recent systematic review of studies evaluating patients with suspected lung cancer looked into the diagnostic sensitivity of various methods of histologic and cytologic tests.¹⁵ Researchers compared the evaluated test results to a reference standard of pathology/histology, definitive cytology, or at least 1-year radiographic follow-up.

Transbronchial needle aspiration showed a sensitivity of 67% (95% CI, 64%–70%) for peripheral lung malignancy of any size; however, only 5 studies met study criteria and their sample sizes varied greatly (n=20 to n=480). Eight studies looking at bronchoscopy (including brush or biopsy) for peripheral lung lesions <2 cm in diameter yielded a sensitivity of only 33% (95% CI, 28%–38%). In the same systematic review, 61 studies of transthoracic needle aspiration for localized pulmonary lesions of any size had a pooled sensitivity of 90% (95% CI, 88%–92%). The prevalence of malignancy in the studies ranged from 0.58 to 0.93.¹⁵ Factors affecting heterogeneity between studies included the wide range in study dates, imaging technology used, and study sizes.

What test is most cost-effective?

CT appears cost-effective when the pretest probability of malignancy is <90%; therefore, consider it on virtually all new cases of solitary pulmonary nodules.¹ Also, when CT and pretest risk-assessments are discordant (eg, a patient has a low pretest probability of malignancy but his CT is suggestive of malignancy), the FDG PET scan is the most economically feasible at less than $20,000 per quality-adjusted life year.

Recommendations from others

The American College of Chest Physicians (ACCP)² suggests pursuing no further evaluation if a nodule is unchanged for >2 years or has benign central calcifications. They recommend that physicians perform CT on every patient with a new nodule to characterize the nodule, its location, and the mediastinum. They do not recommend PET scans for nodules <1 cm. Patients who are marginal surgical candidates and have a negative PET scan should have a repeat CT scan in 3 months; serial CTs at 3, 6, 12, and 24 months are suggested, too, if prior chest radiographs are negative.

The ACCP states that transthoracic needle aspiration is not indicated in surgical candidates unless they decline surgery; then transthoracic needle aspiration or a transbronchial approach are the preferred procedure. Transthoracic needle aspiration may also be useful in establishing a diagnosis for patients who are not surgical candidates or who have a high surgical risk.

ACCP expert consensus favors the reference standard of video-assisted thoracoscopic surgery with wedge resection as the ideal method for obtaining tissue diagnosis in consenting, operable patients with solitary pulmonary nodules. Objective evidence is lacking on follow-up monitoring methods for patients with a nodule who do not have a tissue diagnosis and observation alone is chosen. ACCP expert consensus favors a 2-year follow-up with CT scanning at 3, 6, 12, and 24 months to monitor for nodule growth.²

Evidence summary

A solitary pulmonary nodule, or “coin lesion,” is an intraparenchymal finding on chest radiograph or CT that is less than 3 to 4 cm in diameter and not associated with atelectasis or adenopathy. Malignancy rates range from 15% to 75%, depending on the population studied.¹ Although early detection of malignancy portends a major improvement in survival (up to 75% at 5 years following surgical resection of stage IA disease), most lung cancers progress asymptomatically until quite advanced.²

The presumed benign nature of lesions that are either unchanged over 2 years or have central calcifications is based on 3 retrospective studies from the 1950s.^3-6 However, these should not be considered absolutes. A recent study revisiting the original data calculated the predictive value of benign nature based on no growth to be only 65% (95% confidence interval [CI], 47%–83%).⁷ Also, a study assessing the accuracy of radiologists’ assessment of calcification in solitary pulmonary nodules compared with thin-section CT found that 7% of “definitely calcified” nodules on chest radiograph lacked calcification on thin-section CT.⁸

Which clinical variables best predict malignancy?

The best available clinical decision rule was derived and validated from a single split population of patients with solitary pulmonary nodules.⁹ The outcome variable was defined as malignancy based on histologic tissue analysis or benignity by radiographic stability or resolution over 2 years. The authors did not report whether those determining outcomes and predictors were appropriately blinded.

The authors found that 3 clinical variables (age, smoking history, and cancer history) plus 3 radiographic variables (diameter, spiculation, and nodule location in the upper lobes) were independent predictors of malignancy. An online calculator using this prediction model is available at www.chestx-ray.com/SPN/ SPNProb.html.¹⁰

CT or PET?

Three comparative studies observed 8 to 12 radiologists’ readings of high-resolution CT images of 28 to 56 patients with solitary pulmonary nodules (established diagnoses by either histology or stability over time).^11-13 Approximately half the nodules represented malignant lesions.

Radiologists assigned a level of confidence to their assessment of each case as benign or malignant. At a minimum, they were informed of each patient’s age and gender, and in 2 studies they also knew other information, such as the patient’s smoking and cancer histories. The study showed that the radiologists would have correctly diagnosed a pair of solitary pulmonary nodule cases, one malignant and one benign, between 75% and 83% of the time. Conversely, 17% to 25% of the time they would have diagnosed the case pair incorrectly.

A meta-analysis of 40 studies of FDG PET scanning for solitary pulmonary nodules yielded a maximum joint sensitivity and specificity of 90% (95% CI, 86.4%– 92.7%).¹⁴ The methodological quality of studies included in the meta-analysis was fair, with small sample sizes (inclusion criteria were for a minimum of 10 patients with pulmonary nodules and malignant prevalence of at least 0.5); masking was frequently incomplete.

Sensitivity of histologic/cytologic tests varies

A recent systematic review of studies evaluating patients with suspected lung cancer looked into the diagnostic sensitivity of various methods of histologic and cytologic tests.¹⁵ Researchers compared the evaluated test results to a reference standard of pathology/histology, definitive cytology, or at least 1-year radiographic follow-up.

Transbronchial needle aspiration showed a sensitivity of 67% (95% CI, 64%–70%) for peripheral lung malignancy of any size; however, only 5 studies met study criteria and their sample sizes varied greatly (n=20 to n=480). Eight studies looking at bronchoscopy (including brush or biopsy) for peripheral lung lesions <2 cm in diameter yielded a sensitivity of only 33% (95% CI, 28%–38%). In the same systematic review, 61 studies of transthoracic needle aspiration for localized pulmonary lesions of any size had a pooled sensitivity of 90% (95% CI, 88%–92%). The prevalence of malignancy in the studies ranged from 0.58 to 0.93.¹⁵ Factors affecting heterogeneity between studies included the wide range in study dates, imaging technology used, and study sizes.

What test is most cost-effective?

CT appears cost-effective when the pretest probability of malignancy is <90%; therefore, consider it on virtually all new cases of solitary pulmonary nodules.¹ Also, when CT and pretest risk-assessments are discordant (eg, a patient has a low pretest probability of malignancy but his CT is suggestive of malignancy), the FDG PET scan is the most economically feasible at less than $20,000 per quality-adjusted life year.

Recommendations from others

The American College of Chest Physicians (ACCP)² suggests pursuing no further evaluation if a nodule is unchanged for >2 years or has benign central calcifications. They recommend that physicians perform CT on every patient with a new nodule to characterize the nodule, its location, and the mediastinum. They do not recommend PET scans for nodules <1 cm. Patients who are marginal surgical candidates and have a negative PET scan should have a repeat CT scan in 3 months; serial CTs at 3, 6, 12, and 24 months are suggested, too, if prior chest radiographs are negative.

The ACCP states that transthoracic needle aspiration is not indicated in surgical candidates unless they decline surgery; then transthoracic needle aspiration or a transbronchial approach are the preferred procedure. Transthoracic needle aspiration may also be useful in establishing a diagnosis for patients who are not surgical candidates or who have a high surgical risk.

ACCP expert consensus favors the reference standard of video-assisted thoracoscopic surgery with wedge resection as the ideal method for obtaining tissue diagnosis in consenting, operable patients with solitary pulmonary nodules. Objective evidence is lacking on follow-up monitoring methods for patients with a nodule who do not have a tissue diagnosis and observation alone is chosen. ACCP expert consensus favors a 2-year follow-up with CT scanning at 3, 6, 12, and 24 months to monitor for nodule growth.²

Social risks from genetic testing are a major public policy concern in medical genetics and genetics research.^1-3 These concerns are focused mainly on health insurance, as insurers have an incentive to avoid those clients it will be costly for them to insure. This concern also applies to employers, who pay for most private health care in the US.

Since the beginnings of the Human Genome Project in 1990, legislation has been proposed to head off possible genetic discrimination. The Health Insurance Portability and Accountability Act of 1996 (HIPAA) prohibited most insurers from considering genetic risks as preexisting conditions. Since the 1990s, all but 2 states have passed laws that limit genetic discrimination in health insurance; 34 states ban or limit genetic discrimination by employers.⁴ In addition, a piece of federal legislation is pending. The Genetic Information Nondiscrimination Act, intended to prohibit health insurance and employment discrimination on the basis of genetic information, passed the House in April 2007 but, at press time, had not been voted on by the Senate.

Because state laws usually do not apply to life or disability insurance, and their applicability to employment is inconsistent, the potential for discrimination problems is still of concern to patients with costly conditions that may be uncovered by genetic screening.

Hereditary hemochromatosis—early detection is key

Members of the health care community have expressed concern that genetic discrimination will hamper efforts to detect and mitigate hereditary hemochromatosis, a relatively common condition marked by iron overload, which can lead to irreversible organ damage and related health problems.^5-9 Hereditary hemochromatosis is an autosomal recessive condition that typically is associated with 2 copies of the C282Y mutation in the HFE gene on chromosome 6, although the H63D HFE mutation is also associated with hereditary hemochromatosis in some cases.^10,11 If detected early, health problems from iron overload can be prevented by periodic phlebotomy. Not surprisingly, then, there is public health interest in the screening and detection of hemochromatosis or iron overload before symptoms arise.^9,12-14

There are documented instances, however, of healthy people who experienced discrimination in insurance or employment based on having a phenotype or genotype associated with hereditary hemochromatosis.^8,15,16 Alper and colleagues noted that “[i]n a general screening program of 100 million people…even if only 5% of these people [who screen positive for hereditary hemochromatosis] encounter discrimination, this amounts to approximately 28,000 people.”⁸ Alper’s speculation has not been tested because no large-scale studies of discrimination resulting from screening the general population have been conducted.

This study looks to identify possible social risks from genetic screening by surveying a large and diverse racial/ethnic primary care population participating in a study of iron overload and hereditary hemochromatosis in the US and Canada. We have sought to determine the number and types of problems associated with insurance and employment 1 year after screening and clinical examination for relevant hereditary hemochromatosis phenotypes and genotypes.

Methods

The data for this analysis come from the Hemochromatosis and Iron Overload Screening (HEIRS) Study. A full description of the study’s rationale, design, and methods has been published elsewhere.¹⁷

Identifying the subjects

The research team screened a multiracial/ethnic sample of 101,168 primary care patients, 25 years of age and older, over a 2-year period beginning in March 2001. We recruited participants in approximately equal numbers at 5 field centers, 4 in the US and 1 in Canada. We chose study sites and recruitment goals to produce a sample with about 50% non-Caucasians. A central laboratory assayed blood samples for transferrin saturation, serum ferritin, and HFE C282Y and H63D mutations. A comprehensive clinical evaluation was offered to all C282Y homozygotes and to all non-C282Y homozygotes who met study thresholds for elevated transferrin saturation and serum ferritin iron measures.¹⁷

We identified a total of 2253 such participants. Of these, 1677 (74%) participated in the clinical exam, which assessed body iron stores and attempted to distinguish between primary and secondary causes of iron overload. Clinical exam results were provided to the participants—and, if they consented, to their primary care providers. We then mailed an extended survey to all 1677 participants assessing various psychological and behavioral issues related to genetic screening and testing; 1154 responded (68.8%).

Follow-up 1 year later

One year after the clinical exam, we resurveyed these 1154 respondents and asked additional questions about any problems with employment or insurance (health, life, or disability) in the past year “related to hemochromatosis or iron overload.” Participants who had problems were contacted by telephone for follow-up questions about the nature and circumstances of the problem.

For comparison with clinical exam participants, we surveyed a stratified random sample of 939 screening participants in a similar fashion 1 year following screening. These participants were not eligible for a clinical exam but they had inconclusive screening results, indicating possible elevated risk of hereditary hemochromatosis/iron overload, such as HFE genotypes other than C282Y homozygosity, or lesser elevations of transferrin saturation or serum ferritin.¹⁸ Also surveyed at 1 year were 803 controls (469 of whom had a clinical exam and 334 who did not), who were randomly selected following the age distribution of the other participants. All controls had no known HFE genotypes associated with iron overload and had transferrin saturation and serum ferritin values below study-defined thresholds for risk of iron overload.

Results

Overall, 832 clinical exam participants (72.1%) responded among the 1154 who were surveyed 1 year after their exam. Sample characteristics are shown in TABLE 1.

Few discrimination problems were found

Twenty-four (2.8%) reported they had 1 or more problems with employment or insurance that they believed were related to hereditary hemochromatosis/iron overload. Researchers made follow-up phone contact with 17 of these (70.8%), but 7 were lost to follow-up. Only 3 of these 17 participants verified problems that were possibly related to the participation in the HEIRS Study. Among the comparison group respondents (controls and inconclusive screening results), 6 (0.5%) reported a problem with employment or insurance. However, after we made contact with 5 of the 6 respondents, we were unable to verify any of the 5 cases.

All 3 verified reports of problems came from participants who received a clinical exam based on elevated transferrin saturation and serum ferritin levels, rather than based on their HFE genotypes. There were no verified reports from any of the 153 newly identified C282Y homozygotes who responded (out of a total of 252 such participants in the HEIRS Study). Details of the 3 verified problems are shown in TABLE 2.

One person with primary iron overload was denied long-term care insurance. Two people with elevated transferrin saturation and serum ferritin levels of undetermined cause reported having problems with life insurance. One was charged an increased rate, and the second one was initially refused insurance but was later covered after a physician explained that serum iron measures had returned to normal. There were no verified reports of problems with health insurance or employment. (None of the 7 participants who reported problems but who were lost to further follow-up were newly identified C282Y homozygotes, nor did any have C282Y or H63D HFE mutations.)

TABLE 1
Characteristics of all who responded to follow-up

Discussion

Legal definitions and jurisdictions: Their role in discrimination

In the present study, all 3 participants who reported problems received a clinical examination based on elevated blood iron measures (transferrin saturation and serum ferritin). We verified no problems among participants with any C282Y or H63D HFE genotypes (including both homozygotes or heterozygotes). Therefore, under prevailing definitions, the reported problems do not appear to constitute “genetic discrimination.”¹⁵ Furthermore, we received no verified reports of problems with employment or health insurance, for which legal protections are much stronger than for other types of potential discrimination.¹⁹

Of note, though: None of our study jurisdictions had legal protections against genetic discrimination by life insurers or long-term-care insurers, and these are the 2 areas where our participants encountered problems. This suggests that existing legal protections may be somewhat effective, although it is also consistent with the argument that such protections address potential problems that are rare or nonexistent.¹

Examining the individual case reports (TABLE 2), insurance was denied outright in just one case, involving long-term care insurance, and that participant had health problems unrelated to iron overload that could have contributed to the denial. In the 2 life insurance cases, one person was insured with the help of a physician’s letter, and the other person obtained insurance with an increased premium.

TABLE 2
Characteristics of the 3 participants reporting insurance problems

Limitations of this study

Because this study had only a 1-year follow-up period, it provides no basis for determining the long-term prevalence of insurance or employment problems. However, it appears that after 1 year, the extent of these problems is very small. No verified problems were reported by newly identified C282Y homozygotes or by participants with any other C282Y or H63D HFE genotypes.

Our findings contrast with the 20% prevalence of discrimination among hemochromatosis patients reported by Shaheen et al.¹⁶ Their study, however, evaluated subjects who had been diagnosed in routine medical care to have hemochromatosis an average of 4.5 years before discrimination evaluation, and who were under treatment for iron overload.

Verification contacts in our study were made with those reporting a problem, which may have resulted in some underreporting among respondents. Also, we did not determine which participants sought or changed insurance during the study period, so we do not know the exposure rate to potential discrimination. Nevertheless, these findings provide some empirical basis for informing clinicians, researchers, patients, and Institution Review Boards that the risk of insurance or employment problems following genetic screening for hemochromatosis appears to be quite small.

Funding/Support

This study is supported by contracts N01-HC05185 (University of Minnesota), N01-HC05186 (Howard University), N01-HC05188 (University of Alabama at Birmingham), N01-HC05189 (Kaiser Permanente Center for Health Research), N01-HC05190 (University of California, Irvine), N01-HC05191 (London Health Sciences Centre), and N01-HC05192 (Wake Forest University).

Additional support was provided by the Howard University General Clinical Research Center (GCRC) grant, M01-RR 10284, and the UCSD/UCI Satellite GCRC grant, M01-RR 00827 (University of California, Irvine), sponsored by the National Center for Research Resources, National Institutes of Health. Additional support was provided by the University of Alabama at Birmingham General Clinical Research Center (GCRC) grant M01-RR 00032, Southern Iron Disorders Center, Howard University GCRC grant M01-RR 10284, Howard University Research Scientist Award UH1-HL03679-05 from the National Heart, Lung, and Blood Institute and the Office of Research on Minority Health; and grant UC Irvine M01 RR 000827 from the General Clinical Research Centers Program of the National Center for Research Resources National Institutes of Health. These funding organizations participated in the study’s design and conduct, and they approved publication of this paper, but they did not directly participate in the collection, management, analysis, or interpretation of the data or the preparation of this manuscript.

All authors had full access to all of the data in the study and they take responsibility for the integrity of the data and the accuracy of the data analysis.

The HEIRS Study investigators are listed at www.heirs-study.org/PP_Policy.htm.

CorrespondenceMark A. Hall, JD, Wake Forest University Medical School, Department of Public Health Sciences, 2000 W. 1st Street, Winston-Salem NC 27157-1063; mhall@wfubmc.edu.

Social risks from genetic testing are a major public policy concern in medical genetics and genetics research.^1-3 These concerns are focused mainly on health insurance, as insurers have an incentive to avoid those clients it will be costly for them to insure. This concern also applies to employers, who pay for most private health care in the US.

Since the beginnings of the Human Genome Project in 1990, legislation has been proposed to head off possible genetic discrimination. The Health Insurance Portability and Accountability Act of 1996 (HIPAA) prohibited most insurers from considering genetic risks as preexisting conditions. Since the 1990s, all but 2 states have passed laws that limit genetic discrimination in health insurance; 34 states ban or limit genetic discrimination by employers.⁴ In addition, a piece of federal legislation is pending. The Genetic Information Nondiscrimination Act, intended to prohibit health insurance and employment discrimination on the basis of genetic information, passed the House in April 2007 but, at press time, had not been voted on by the Senate.

Because state laws usually do not apply to life or disability insurance, and their applicability to employment is inconsistent, the potential for discrimination problems is still of concern to patients with costly conditions that may be uncovered by genetic screening.

Hereditary hemochromatosis—early detection is key

Members of the health care community have expressed concern that genetic discrimination will hamper efforts to detect and mitigate hereditary hemochromatosis, a relatively common condition marked by iron overload, which can lead to irreversible organ damage and related health problems.^5-9 Hereditary hemochromatosis is an autosomal recessive condition that typically is associated with 2 copies of the C282Y mutation in the HFE gene on chromosome 6, although the H63D HFE mutation is also associated with hereditary hemochromatosis in some cases.^10,11 If detected early, health problems from iron overload can be prevented by periodic phlebotomy. Not surprisingly, then, there is public health interest in the screening and detection of hemochromatosis or iron overload before symptoms arise.^9,12-14

There are documented instances, however, of healthy people who experienced discrimination in insurance or employment based on having a phenotype or genotype associated with hereditary hemochromatosis.^8,15,16 Alper and colleagues noted that “[i]n a general screening program of 100 million people…even if only 5% of these people [who screen positive for hereditary hemochromatosis] encounter discrimination, this amounts to approximately 28,000 people.”⁸ Alper’s speculation has not been tested because no large-scale studies of discrimination resulting from screening the general population have been conducted.

This study looks to identify possible social risks from genetic screening by surveying a large and diverse racial/ethnic primary care population participating in a study of iron overload and hereditary hemochromatosis in the US and Canada. We have sought to determine the number and types of problems associated with insurance and employment 1 year after screening and clinical examination for relevant hereditary hemochromatosis phenotypes and genotypes.

Methods

The data for this analysis come from the Hemochromatosis and Iron Overload Screening (HEIRS) Study. A full description of the study’s rationale, design, and methods has been published elsewhere.¹⁷

Identifying the subjects

The research team screened a multiracial/ethnic sample of 101,168 primary care patients, 25 years of age and older, over a 2-year period beginning in March 2001. We recruited participants in approximately equal numbers at 5 field centers, 4 in the US and 1 in Canada. We chose study sites and recruitment goals to produce a sample with about 50% non-Caucasians. A central laboratory assayed blood samples for transferrin saturation, serum ferritin, and HFE C282Y and H63D mutations. A comprehensive clinical evaluation was offered to all C282Y homozygotes and to all non-C282Y homozygotes who met study thresholds for elevated transferrin saturation and serum ferritin iron measures.¹⁷

We identified a total of 2253 such participants. Of these, 1677 (74%) participated in the clinical exam, which assessed body iron stores and attempted to distinguish between primary and secondary causes of iron overload. Clinical exam results were provided to the participants—and, if they consented, to their primary care providers. We then mailed an extended survey to all 1677 participants assessing various psychological and behavioral issues related to genetic screening and testing; 1154 responded (68.8%).

Follow-up 1 year later

One year after the clinical exam, we resurveyed these 1154 respondents and asked additional questions about any problems with employment or insurance (health, life, or disability) in the past year “related to hemochromatosis or iron overload.” Participants who had problems were contacted by telephone for follow-up questions about the nature and circumstances of the problem.

For comparison with clinical exam participants, we surveyed a stratified random sample of 939 screening participants in a similar fashion 1 year following screening. These participants were not eligible for a clinical exam but they had inconclusive screening results, indicating possible elevated risk of hereditary hemochromatosis/iron overload, such as HFE genotypes other than C282Y homozygosity, or lesser elevations of transferrin saturation or serum ferritin.¹⁸ Also surveyed at 1 year were 803 controls (469 of whom had a clinical exam and 334 who did not), who were randomly selected following the age distribution of the other participants. All controls had no known HFE genotypes associated with iron overload and had transferrin saturation and serum ferritin values below study-defined thresholds for risk of iron overload.

Results

Overall, 832 clinical exam participants (72.1%) responded among the 1154 who were surveyed 1 year after their exam. Sample characteristics are shown in TABLE 1.

Few discrimination problems were found

Twenty-four (2.8%) reported they had 1 or more problems with employment or insurance that they believed were related to hereditary hemochromatosis/iron overload. Researchers made follow-up phone contact with 17 of these (70.8%), but 7 were lost to follow-up. Only 3 of these 17 participants verified problems that were possibly related to the participation in the HEIRS Study. Among the comparison group respondents (controls and inconclusive screening results), 6 (0.5%) reported a problem with employment or insurance. However, after we made contact with 5 of the 6 respondents, we were unable to verify any of the 5 cases.

All 3 verified reports of problems came from participants who received a clinical exam based on elevated transferrin saturation and serum ferritin levels, rather than based on their HFE genotypes. There were no verified reports from any of the 153 newly identified C282Y homozygotes who responded (out of a total of 252 such participants in the HEIRS Study). Details of the 3 verified problems are shown in TABLE 2.

One person with primary iron overload was denied long-term care insurance. Two people with elevated transferrin saturation and serum ferritin levels of undetermined cause reported having problems with life insurance. One was charged an increased rate, and the second one was initially refused insurance but was later covered after a physician explained that serum iron measures had returned to normal. There were no verified reports of problems with health insurance or employment. (None of the 7 participants who reported problems but who were lost to further follow-up were newly identified C282Y homozygotes, nor did any have C282Y or H63D HFE mutations.)

TABLE 1
Characteristics of all who responded to follow-up

Discussion

Legal definitions and jurisdictions: Their role in discrimination

In the present study, all 3 participants who reported problems received a clinical examination based on elevated blood iron measures (transferrin saturation and serum ferritin). We verified no problems among participants with any C282Y or H63D HFE genotypes (including both homozygotes or heterozygotes). Therefore, under prevailing definitions, the reported problems do not appear to constitute “genetic discrimination.”¹⁵ Furthermore, we received no verified reports of problems with employment or health insurance, for which legal protections are much stronger than for other types of potential discrimination.¹⁹

Of note, though: None of our study jurisdictions had legal protections against genetic discrimination by life insurers or long-term-care insurers, and these are the 2 areas where our participants encountered problems. This suggests that existing legal protections may be somewhat effective, although it is also consistent with the argument that such protections address potential problems that are rare or nonexistent.¹

Examining the individual case reports (TABLE 2), insurance was denied outright in just one case, involving long-term care insurance, and that participant had health problems unrelated to iron overload that could have contributed to the denial. In the 2 life insurance cases, one person was insured with the help of a physician’s letter, and the other person obtained insurance with an increased premium.

TABLE 2
Characteristics of the 3 participants reporting insurance problems

Limitations of this study

Because this study had only a 1-year follow-up period, it provides no basis for determining the long-term prevalence of insurance or employment problems. However, it appears that after 1 year, the extent of these problems is very small. No verified problems were reported by newly identified C282Y homozygotes or by participants with any other C282Y or H63D HFE genotypes.

Our findings contrast with the 20% prevalence of discrimination among hemochromatosis patients reported by Shaheen et al.¹⁶ Their study, however, evaluated subjects who had been diagnosed in routine medical care to have hemochromatosis an average of 4.5 years before discrimination evaluation, and who were under treatment for iron overload.

Verification contacts in our study were made with those reporting a problem, which may have resulted in some underreporting among respondents. Also, we did not determine which participants sought or changed insurance during the study period, so we do not know the exposure rate to potential discrimination. Nevertheless, these findings provide some empirical basis for informing clinicians, researchers, patients, and Institution Review Boards that the risk of insurance or employment problems following genetic screening for hemochromatosis appears to be quite small.

Funding/Support

This study is supported by contracts N01-HC05185 (University of Minnesota), N01-HC05186 (Howard University), N01-HC05188 (University of Alabama at Birmingham), N01-HC05189 (Kaiser Permanente Center for Health Research), N01-HC05190 (University of California, Irvine), N01-HC05191 (London Health Sciences Centre), and N01-HC05192 (Wake Forest University).

Additional support was provided by the Howard University General Clinical Research Center (GCRC) grant, M01-RR 10284, and the UCSD/UCI Satellite GCRC grant, M01-RR 00827 (University of California, Irvine), sponsored by the National Center for Research Resources, National Institutes of Health. Additional support was provided by the University of Alabama at Birmingham General Clinical Research Center (GCRC) grant M01-RR 00032, Southern Iron Disorders Center, Howard University GCRC grant M01-RR 10284, Howard University Research Scientist Award UH1-HL03679-05 from the National Heart, Lung, and Blood Institute and the Office of Research on Minority Health; and grant UC Irvine M01 RR 000827 from the General Clinical Research Centers Program of the National Center for Research Resources National Institutes of Health. These funding organizations participated in the study’s design and conduct, and they approved publication of this paper, but they did not directly participate in the collection, management, analysis, or interpretation of the data or the preparation of this manuscript.

All authors had full access to all of the data in the study and they take responsibility for the integrity of the data and the accuracy of the data analysis.

The HEIRS Study investigators are listed at www.heirs-study.org/PP_Policy.htm.

CorrespondenceMark A. Hall, JD, Wake Forest University Medical School, Department of Public Health Sciences, 2000 W. 1st Street, Winston-Salem NC 27157-1063; mhall@wfubmc.edu.

Social risks from genetic testing are a major public policy concern in medical genetics and genetics research.^1-3 These concerns are focused mainly on health insurance, as insurers have an incentive to avoid those clients it will be costly for them to insure. This concern also applies to employers, who pay for most private health care in the US.

Since the beginnings of the Human Genome Project in 1990, legislation has been proposed to head off possible genetic discrimination. The Health Insurance Portability and Accountability Act of 1996 (HIPAA) prohibited most insurers from considering genetic risks as preexisting conditions. Since the 1990s, all but 2 states have passed laws that limit genetic discrimination in health insurance; 34 states ban or limit genetic discrimination by employers.⁴ In addition, a piece of federal legislation is pending. The Genetic Information Nondiscrimination Act, intended to prohibit health insurance and employment discrimination on the basis of genetic information, passed the House in April 2007 but, at press time, had not been voted on by the Senate.

Because state laws usually do not apply to life or disability insurance, and their applicability to employment is inconsistent, the potential for discrimination problems is still of concern to patients with costly conditions that may be uncovered by genetic screening.

Hereditary hemochromatosis—early detection is key

Members of the health care community have expressed concern that genetic discrimination will hamper efforts to detect and mitigate hereditary hemochromatosis, a relatively common condition marked by iron overload, which can lead to irreversible organ damage and related health problems.^5-9 Hereditary hemochromatosis is an autosomal recessive condition that typically is associated with 2 copies of the C282Y mutation in the HFE gene on chromosome 6, although the H63D HFE mutation is also associated with hereditary hemochromatosis in some cases.^10,11 If detected early, health problems from iron overload can be prevented by periodic phlebotomy. Not surprisingly, then, there is public health interest in the screening and detection of hemochromatosis or iron overload before symptoms arise.^9,12-14

There are documented instances, however, of healthy people who experienced discrimination in insurance or employment based on having a phenotype or genotype associated with hereditary hemochromatosis.^8,15,16 Alper and colleagues noted that “[i]n a general screening program of 100 million people…even if only 5% of these people [who screen positive for hereditary hemochromatosis] encounter discrimination, this amounts to approximately 28,000 people.”⁸ Alper’s speculation has not been tested because no large-scale studies of discrimination resulting from screening the general population have been conducted.

This study looks to identify possible social risks from genetic screening by surveying a large and diverse racial/ethnic primary care population participating in a study of iron overload and hereditary hemochromatosis in the US and Canada. We have sought to determine the number and types of problems associated with insurance and employment 1 year after screening and clinical examination for relevant hereditary hemochromatosis phenotypes and genotypes.

Methods

The data for this analysis come from the Hemochromatosis and Iron Overload Screening (HEIRS) Study. A full description of the study’s rationale, design, and methods has been published elsewhere.¹⁷

Identifying the subjects

The research team screened a multiracial/ethnic sample of 101,168 primary care patients, 25 years of age and older, over a 2-year period beginning in March 2001. We recruited participants in approximately equal numbers at 5 field centers, 4 in the US and 1 in Canada. We chose study sites and recruitment goals to produce a sample with about 50% non-Caucasians. A central laboratory assayed blood samples for transferrin saturation, serum ferritin, and HFE C282Y and H63D mutations. A comprehensive clinical evaluation was offered to all C282Y homozygotes and to all non-C282Y homozygotes who met study thresholds for elevated transferrin saturation and serum ferritin iron measures.¹⁷

We identified a total of 2253 such participants. Of these, 1677 (74%) participated in the clinical exam, which assessed body iron stores and attempted to distinguish between primary and secondary causes of iron overload. Clinical exam results were provided to the participants—and, if they consented, to their primary care providers. We then mailed an extended survey to all 1677 participants assessing various psychological and behavioral issues related to genetic screening and testing; 1154 responded (68.8%).

Follow-up 1 year later

One year after the clinical exam, we resurveyed these 1154 respondents and asked additional questions about any problems with employment or insurance (health, life, or disability) in the past year “related to hemochromatosis or iron overload.” Participants who had problems were contacted by telephone for follow-up questions about the nature and circumstances of the problem.

For comparison with clinical exam participants, we surveyed a stratified random sample of 939 screening participants in a similar fashion 1 year following screening. These participants were not eligible for a clinical exam but they had inconclusive screening results, indicating possible elevated risk of hereditary hemochromatosis/iron overload, such as HFE genotypes other than C282Y homozygosity, or lesser elevations of transferrin saturation or serum ferritin.¹⁸ Also surveyed at 1 year were 803 controls (469 of whom had a clinical exam and 334 who did not), who were randomly selected following the age distribution of the other participants. All controls had no known HFE genotypes associated with iron overload and had transferrin saturation and serum ferritin values below study-defined thresholds for risk of iron overload.

Results

Overall, 832 clinical exam participants (72.1%) responded among the 1154 who were surveyed 1 year after their exam. Sample characteristics are shown in TABLE 1.

Few discrimination problems were found

Twenty-four (2.8%) reported they had 1 or more problems with employment or insurance that they believed were related to hereditary hemochromatosis/iron overload. Researchers made follow-up phone contact with 17 of these (70.8%), but 7 were lost to follow-up. Only 3 of these 17 participants verified problems that were possibly related to the participation in the HEIRS Study. Among the comparison group respondents (controls and inconclusive screening results), 6 (0.5%) reported a problem with employment or insurance. However, after we made contact with 5 of the 6 respondents, we were unable to verify any of the 5 cases.

All 3 verified reports of problems came from participants who received a clinical exam based on elevated transferrin saturation and serum ferritin levels, rather than based on their HFE genotypes. There were no verified reports from any of the 153 newly identified C282Y homozygotes who responded (out of a total of 252 such participants in the HEIRS Study). Details of the 3 verified problems are shown in TABLE 2.

One person with primary iron overload was denied long-term care insurance. Two people with elevated transferrin saturation and serum ferritin levels of undetermined cause reported having problems with life insurance. One was charged an increased rate, and the second one was initially refused insurance but was later covered after a physician explained that serum iron measures had returned to normal. There were no verified reports of problems with health insurance or employment. (None of the 7 participants who reported problems but who were lost to further follow-up were newly identified C282Y homozygotes, nor did any have C282Y or H63D HFE mutations.)

TABLE 1
Characteristics of all who responded to follow-up

Discussion

Legal definitions and jurisdictions: Their role in discrimination

In the present study, all 3 participants who reported problems received a clinical examination based on elevated blood iron measures (transferrin saturation and serum ferritin). We verified no problems among participants with any C282Y or H63D HFE genotypes (including both homozygotes or heterozygotes). Therefore, under prevailing definitions, the reported problems do not appear to constitute “genetic discrimination.”¹⁵ Furthermore, we received no verified reports of problems with employment or health insurance, for which legal protections are much stronger than for other types of potential discrimination.¹⁹

Of note, though: None of our study jurisdictions had legal protections against genetic discrimination by life insurers or long-term-care insurers, and these are the 2 areas where our participants encountered problems. This suggests that existing legal protections may be somewhat effective, although it is also consistent with the argument that such protections address potential problems that are rare or nonexistent.¹

Examining the individual case reports (TABLE 2), insurance was denied outright in just one case, involving long-term care insurance, and that participant had health problems unrelated to iron overload that could have contributed to the denial. In the 2 life insurance cases, one person was insured with the help of a physician’s letter, and the other person obtained insurance with an increased premium.

TABLE 2
Characteristics of the 3 participants reporting insurance problems

Limitations of this study

Because this study had only a 1-year follow-up period, it provides no basis for determining the long-term prevalence of insurance or employment problems. However, it appears that after 1 year, the extent of these problems is very small. No verified problems were reported by newly identified C282Y homozygotes or by participants with any other C282Y or H63D HFE genotypes.

Our findings contrast with the 20% prevalence of discrimination among hemochromatosis patients reported by Shaheen et al.¹⁶ Their study, however, evaluated subjects who had been diagnosed in routine medical care to have hemochromatosis an average of 4.5 years before discrimination evaluation, and who were under treatment for iron overload.

Verification contacts in our study were made with those reporting a problem, which may have resulted in some underreporting among respondents. Also, we did not determine which participants sought or changed insurance during the study period, so we do not know the exposure rate to potential discrimination. Nevertheless, these findings provide some empirical basis for informing clinicians, researchers, patients, and Institution Review Boards that the risk of insurance or employment problems following genetic screening for hemochromatosis appears to be quite small.

Funding/Support

This study is supported by contracts N01-HC05185 (University of Minnesota), N01-HC05186 (Howard University), N01-HC05188 (University of Alabama at Birmingham), N01-HC05189 (Kaiser Permanente Center for Health Research), N01-HC05190 (University of California, Irvine), N01-HC05191 (London Health Sciences Centre), and N01-HC05192 (Wake Forest University).

Additional support was provided by the Howard University General Clinical Research Center (GCRC) grant, M01-RR 10284, and the UCSD/UCI Satellite GCRC grant, M01-RR 00827 (University of California, Irvine), sponsored by the National Center for Research Resources, National Institutes of Health. Additional support was provided by the University of Alabama at Birmingham General Clinical Research Center (GCRC) grant M01-RR 00032, Southern Iron Disorders Center, Howard University GCRC grant M01-RR 10284, Howard University Research Scientist Award UH1-HL03679-05 from the National Heart, Lung, and Blood Institute and the Office of Research on Minority Health; and grant UC Irvine M01 RR 000827 from the General Clinical Research Centers Program of the National Center for Research Resources National Institutes of Health. These funding organizations participated in the study’s design and conduct, and they approved publication of this paper, but they did not directly participate in the collection, management, analysis, or interpretation of the data or the preparation of this manuscript.

All authors had full access to all of the data in the study and they take responsibility for the integrity of the data and the accuracy of the data analysis.

The HEIRS Study investigators are listed at www.heirs-study.org/PP_Policy.htm.

CorrespondenceMark A. Hall, JD, Wake Forest University Medical School, Department of Public Health Sciences, 2000 W. 1st Street, Winston-Salem NC 27157-1063; mhall@wfubmc.edu.

Each year, the flu causes an average of about 36,000 excess deaths and over 200,000 hospitalizations in the US.^1,2 Much of this morbidity and mortality is preventable, yet each year, a large proportion of those for whom the vaccine is recommended go unvaccinated (TABLE 1).

TABLE 1
High-risk groups who went unvaccinated with influenza vaccine (2005)

Improving rates among health care workers

The recommendations of the Centers for Disease Control and Prevention (CDC) for the 2007–2008 influenza season include a new recommendation that targets health care worker vaccination rates.³ Because of the low rate of vaccination of health care workers, and the potential impact of higher coverage on both worker and patient safety, the CDC now recommends that the level of vaccination coverage be used as one measure of a facility’s patient safety quality program. The CDC also recommends the implementation of policies to encourage acceptance of the vaccine, such as requiring those caregivers who refuse immunization to sign waivers.

Improving rates among patients

To improve vaccination levels among patients, the CDC recommends:

• using reminder/recall systems
• using standing order programs
• administering the vaccine before and during the influenza season to patients during routine health care visits.

For more on improving vaccination coverage, see “Tips to help improve vaccination rates”.

Offer the vaccine to anyone who wants it

While the groups for whom vaccine is recommended are the same as last year (TABLE 2), this year the CDC is emphasizing the importance of:

• offering the vaccine to anyone who wants to reduce their risk of contracting influenza or transmitting the virus to others.
• continuing to offer vaccine to those susceptible throughout the flu season.

A minor change from last year’s recommendations involves children who are 6 months through 8 years of age who receive only 1 dose of vaccine their first year of vaccination. The CDC now recommends that these children receive 2 doses the next year. If they receive only 1 dose 2 years in a row, the CDC recommends only a single dose annually thereafter.

TABLE 2
Who should receive the influenza vaccine?

The 2 vaccines: How they differ

The same 2 vaccine types are available this year as last: trivalent influenza vaccine (TIV) and live attenuated influenza vaccine (LAIV). The vaccines include the same viral strain antigens and either can be used annually unless contraindicated (TABLE 3).

The major differences between the 2 vaccine types are:

• LAIV is administered as an intranasal spray while TIV requires an intramuscular injection
• LAIV is approved only for healthy people who are 5 to 49 years of age, whereas TIV is approved for anyone over the age of 6 months
• The interval between 2 doses in children under 9 years of age is 4 weeks for TIV and 6 to 10 weeks for LAIV
• LAIV should not be administered to family members or close contacts of those who are immunosuppressed and require a protective environment, while TIV can be used in this situation
• LAIV, being a live virus vaccine, should be administered simultaneously with, or 4 weeks after, the administration of other live virus vaccines. TIV is not a live virus vaccine, and its timing in relation to other live virus vaccines is not an issue.

TABLE 3
Contraindications and precautions for influenza vaccines

Antiviral options remain the same

Once again this year, the CDC does not recommend the use of adamantane antivirals for prophylaxis or treatment of influenza, leaving the 2 neuraminidase inhibitors, oseltamivir (Tamiflu) and zanamivir (Relenza), for these purposes. Treating flu patients with these antivirals shortens the duration of symptoms and may reduce viral shedding.

The earlier the treatment is started, the better the results. There appears to be no—or only minimal—benefit for those with uncomplicated influenza if the treatment is started more than 2 days after the onset of illness.

Consider antiviral prophylaxis for these patients

The CDC recommends that antiviral prophylaxis be considered for those who are susceptible, residing in an area with circulating influenza virus, and who:

• have not been vaccinated or were recently vaccinated (since it takes 2 weeks for immunity to develop after vaccination)
• are unvaccinated and providing care for high-risk individuals
• have a contraindication to the vaccine
• have immune deficiencies and may not respond adequately to the vaccine.

The CDC also recommends prophylaxis for all residents and staff in a long-term care facility where influenza is circulating, without regard to vaccine status. More complete information on indications, dose and duration of antivirals for prophylaxis, and treatment can be found in this year’s CDC recommendations.³

Another flu season approaches

The good news for the coming year is that the government expects that the supply of vaccine will exceed 100 million doses. This should be sufficient, unless unforeseen production problems arise.

Each year millions of doses of influenza vaccine go unused and are discarded. By following the CDC’s recommendations, and those of the Task Force on Community Preventive Services⁴ (top left), each of us can improve vaccination coverage in our area and minimize the number of hospitalizations and deaths from the flu.

Correspondence
Doug Campos-Outcalt, MD, MPA, 550 e. van buren, Phoenix, AZ 85004; dougco@u.arizona.edu.

Each year, the flu causes an average of about 36,000 excess deaths and over 200,000 hospitalizations in the US.^1,2 Much of this morbidity and mortality is preventable, yet each year, a large proportion of those for whom the vaccine is recommended go unvaccinated (TABLE 1).

TABLE 1
High-risk groups who went unvaccinated with influenza vaccine (2005)

Improving rates among health care workers

The recommendations of the Centers for Disease Control and Prevention (CDC) for the 2007–2008 influenza season include a new recommendation that targets health care worker vaccination rates.³ Because of the low rate of vaccination of health care workers, and the potential impact of higher coverage on both worker and patient safety, the CDC now recommends that the level of vaccination coverage be used as one measure of a facility’s patient safety quality program. The CDC also recommends the implementation of policies to encourage acceptance of the vaccine, such as requiring those caregivers who refuse immunization to sign waivers.

Improving rates among patients

To improve vaccination levels among patients, the CDC recommends:

• using reminder/recall systems
• using standing order programs
• administering the vaccine before and during the influenza season to patients during routine health care visits.

For more on improving vaccination coverage, see “Tips to help improve vaccination rates”.

Offer the vaccine to anyone who wants it

While the groups for whom vaccine is recommended are the same as last year (TABLE 2), this year the CDC is emphasizing the importance of:

• offering the vaccine to anyone who wants to reduce their risk of contracting influenza or transmitting the virus to others.
• continuing to offer vaccine to those susceptible throughout the flu season.

A minor change from last year’s recommendations involves children who are 6 months through 8 years of age who receive only 1 dose of vaccine their first year of vaccination. The CDC now recommends that these children receive 2 doses the next year. If they receive only 1 dose 2 years in a row, the CDC recommends only a single dose annually thereafter.

TABLE 2
Who should receive the influenza vaccine?

The 2 vaccines: How they differ

The same 2 vaccine types are available this year as last: trivalent influenza vaccine (TIV) and live attenuated influenza vaccine (LAIV). The vaccines include the same viral strain antigens and either can be used annually unless contraindicated (TABLE 3).

The major differences between the 2 vaccine types are:

• LAIV is administered as an intranasal spray while TIV requires an intramuscular injection
• LAIV is approved only for healthy people who are 5 to 49 years of age, whereas TIV is approved for anyone over the age of 6 months
• The interval between 2 doses in children under 9 years of age is 4 weeks for TIV and 6 to 10 weeks for LAIV
• LAIV should not be administered to family members or close contacts of those who are immunosuppressed and require a protective environment, while TIV can be used in this situation
• LAIV, being a live virus vaccine, should be administered simultaneously with, or 4 weeks after, the administration of other live virus vaccines. TIV is not a live virus vaccine, and its timing in relation to other live virus vaccines is not an issue.

TABLE 3
Contraindications and precautions for influenza vaccines

Antiviral options remain the same

Once again this year, the CDC does not recommend the use of adamantane antivirals for prophylaxis or treatment of influenza, leaving the 2 neuraminidase inhibitors, oseltamivir (Tamiflu) and zanamivir (Relenza), for these purposes. Treating flu patients with these antivirals shortens the duration of symptoms and may reduce viral shedding.

The earlier the treatment is started, the better the results. There appears to be no—or only minimal—benefit for those with uncomplicated influenza if the treatment is started more than 2 days after the onset of illness.

Consider antiviral prophylaxis for these patients

The CDC recommends that antiviral prophylaxis be considered for those who are susceptible, residing in an area with circulating influenza virus, and who:

• have not been vaccinated or were recently vaccinated (since it takes 2 weeks for immunity to develop after vaccination)
• are unvaccinated and providing care for high-risk individuals
• have a contraindication to the vaccine
• have immune deficiencies and may not respond adequately to the vaccine.

The CDC also recommends prophylaxis for all residents and staff in a long-term care facility where influenza is circulating, without regard to vaccine status. More complete information on indications, dose and duration of antivirals for prophylaxis, and treatment can be found in this year’s CDC recommendations.³

Another flu season approaches

The good news for the coming year is that the government expects that the supply of vaccine will exceed 100 million doses. This should be sufficient, unless unforeseen production problems arise.

Each year millions of doses of influenza vaccine go unused and are discarded. By following the CDC’s recommendations, and those of the Task Force on Community Preventive Services⁴ (top left), each of us can improve vaccination coverage in our area and minimize the number of hospitalizations and deaths from the flu.

Correspondence
Doug Campos-Outcalt, MD, MPA, 550 e. van buren, Phoenix, AZ 85004; dougco@u.arizona.edu.

Each year, the flu causes an average of about 36,000 excess deaths and over 200,000 hospitalizations in the US.^1,2 Much of this morbidity and mortality is preventable, yet each year, a large proportion of those for whom the vaccine is recommended go unvaccinated (TABLE 1).

TABLE 1
High-risk groups who went unvaccinated with influenza vaccine (2005)

Improving rates among health care workers

The recommendations of the Centers for Disease Control and Prevention (CDC) for the 2007–2008 influenza season include a new recommendation that targets health care worker vaccination rates.³ Because of the low rate of vaccination of health care workers, and the potential impact of higher coverage on both worker and patient safety, the CDC now recommends that the level of vaccination coverage be used as one measure of a facility’s patient safety quality program. The CDC also recommends the implementation of policies to encourage acceptance of the vaccine, such as requiring those caregivers who refuse immunization to sign waivers.

Improving rates among patients

To improve vaccination levels among patients, the CDC recommends:

• using reminder/recall systems
• using standing order programs
• administering the vaccine before and during the influenza season to patients during routine health care visits.

For more on improving vaccination coverage, see “Tips to help improve vaccination rates”.

Offer the vaccine to anyone who wants it

While the groups for whom vaccine is recommended are the same as last year (TABLE 2), this year the CDC is emphasizing the importance of:

• offering the vaccine to anyone who wants to reduce their risk of contracting influenza or transmitting the virus to others.
• continuing to offer vaccine to those susceptible throughout the flu season.

A minor change from last year’s recommendations involves children who are 6 months through 8 years of age who receive only 1 dose of vaccine their first year of vaccination. The CDC now recommends that these children receive 2 doses the next year. If they receive only 1 dose 2 years in a row, the CDC recommends only a single dose annually thereafter.

TABLE 2
Who should receive the influenza vaccine?

The 2 vaccines: How they differ

The same 2 vaccine types are available this year as last: trivalent influenza vaccine (TIV) and live attenuated influenza vaccine (LAIV). The vaccines include the same viral strain antigens and either can be used annually unless contraindicated (TABLE 3).

The major differences between the 2 vaccine types are:

• LAIV is administered as an intranasal spray while TIV requires an intramuscular injection
• LAIV is approved only for healthy people who are 5 to 49 years of age, whereas TIV is approved for anyone over the age of 6 months
• The interval between 2 doses in children under 9 years of age is 4 weeks for TIV and 6 to 10 weeks for LAIV
• LAIV should not be administered to family members or close contacts of those who are immunosuppressed and require a protective environment, while TIV can be used in this situation
• LAIV, being a live virus vaccine, should be administered simultaneously with, or 4 weeks after, the administration of other live virus vaccines. TIV is not a live virus vaccine, and its timing in relation to other live virus vaccines is not an issue.

TABLE 3
Contraindications and precautions for influenza vaccines

Antiviral options remain the same

Once again this year, the CDC does not recommend the use of adamantane antivirals for prophylaxis or treatment of influenza, leaving the 2 neuraminidase inhibitors, oseltamivir (Tamiflu) and zanamivir (Relenza), for these purposes. Treating flu patients with these antivirals shortens the duration of symptoms and may reduce viral shedding.

The earlier the treatment is started, the better the results. There appears to be no—or only minimal—benefit for those with uncomplicated influenza if the treatment is started more than 2 days after the onset of illness.

Consider antiviral prophylaxis for these patients

The CDC recommends that antiviral prophylaxis be considered for those who are susceptible, residing in an area with circulating influenza virus, and who:

• have not been vaccinated or were recently vaccinated (since it takes 2 weeks for immunity to develop after vaccination)
• are unvaccinated and providing care for high-risk individuals
• have a contraindication to the vaccine
• have immune deficiencies and may not respond adequately to the vaccine.

The CDC also recommends prophylaxis for all residents and staff in a long-term care facility where influenza is circulating, without regard to vaccine status. More complete information on indications, dose and duration of antivirals for prophylaxis, and treatment can be found in this year’s CDC recommendations.³

Another flu season approaches

The good news for the coming year is that the government expects that the supply of vaccine will exceed 100 million doses. This should be sufficient, unless unforeseen production problems arise.

Each year millions of doses of influenza vaccine go unused and are discarded. By following the CDC’s recommendations, and those of the Task Force on Community Preventive Services⁴ (top left), each of us can improve vaccination coverage in our area and minimize the number of hospitalizations and deaths from the flu.

Correspondence
Doug Campos-Outcalt, MD, MPA, 550 e. van buren, Phoenix, AZ 85004; dougco@u.arizona.edu.