User login
The role of psychiatric APRNs
In Dr. Mary Moller’s Guest Editorial “Advancing the role of advanced practice psychiatric nurses in today’s psychiatric workforce” (
Dr. Moller cited a source from the Federal Trade Commission1 that encourages the autonomous practice of APRNs to increase competition. This again implies the false equivalency between physicians and APRNs. Competition implies that the players are providing the same service. If, as nurse practitioners argue, they practice “nursing,” then they are not practicing “medicine.” Physicians and APRNs do not have the same background. Although both are charged with the care of patients, nursing is not medicine, nor should it be. Both are important and needed, but nursing was never designed to be an autonomous practice. According to the American Association of Colleges of Nursing, “Nursing and medicine are distinct health disciplines that prepare clinicians to assume different roles and meet different practice expectations.”2 In fact, the curriculum and requirements to become an APRN vary depending on the program, and some programs do not even require a BSN.3 There are online programs available for earning an APRN degree. Additionally, APRNs are only required to have 500 to 700 total hours of patient care,4 compared with the >10,000 hours physicians have once they have finished a 3-year residency, which when combined with their education amounts to >20,000 hours.5 This doesn’t account for those who have longer residencies or fellowships to further specialize in their area of training.
Dr. Moller’s main argument is that there is a dire shortage of psychiatrists and that the only way to meet this need for more providers is to make APRNs autonomous. However, no data indicate that autonomous practice of mid-level providers leads to an influx of these providers in rural areas, where the need would be greatest. Although current data on this are quite sparse, some studies indicate that the majority practice in urban areas, even in states with independent practice authority.6,7 Dr. Moller cites a source that only reviewed home zip codes of psychiatric APRNs but did not include zip codes of employment.8 Only 13% of psychiatric APRNs live in rural areas across the United States. Therefore, it is a false assertion to state that these APRNs are found primarily in rural and less populated urban areas. It is also false to imply and assume that these APRNs practice in the rural areas.
In 2017, there were 43,157 registered physician applications, with 35,969 active applications for 31,757 residency positions in the United States, and at least 11,400 medical school graduates were unmatched.9 Imagine how much more we could serve our patients by matching these graduates, whose training far surpasses that of a mid-level provider. The Resident Physician Shortage Reduction Act of 2017 aims to address this problem by increasing Medicare-funded graduate medical education (GME) residency programs in the United States.10 We can make a difference by contacting our members of Congress to encourage them to support this bill. In addition, the AMA is advocating to save funding for GME and provides an easy-to-use Web site (https://savegme.org/take-action) to contact your legislators directly to show your support for GME.
Nurse practitioners have tremendous value when their role is a part of a team; however, they should not practice without supervision, and physicians who supervise them absolutely should be providing adequate supervision. I applaud the APA and the AMA for standing up for the practice of medicine and for our patients. I hope that they continue to do so, and I encourage them to increase their efforts.
Laura Kendall, MD
Assistant Professor of Clinical Psychiatry
Department of Psychiatry and Behavioral Sciences
Keck School of Medicine
University of Southern California
Los Angeles, California
References
1. Koslov T; Office of Policy Planning. The doctor (or nurse practitioner) will see you now: competition and the regulation of advanced practice nurses. Federal Trade Commission. https://www.ftc.gov/news-events/blogs/competition-matters/2014/03/doctor-or-nurse-practitioner-will-see-you-now. Published March 7, 2014. Accessed July 26, 2017.
2. American Association of Colleges of Nursing. DNP talking points. http://www.aacnnursing.org/DNP/about/talking-points. Updated July, 2014. Accessed August 12, 2017.
3. Keyes L. MSN without a BSN? MastersInNursing.com. https://www.mastersinnursing.com/msn-without-a-bsn. Accessed August 12, 2017.
4. Iglehart JK. Expanding the role of advanced nurse practitioners—risks and rewards. New Engl J Med. 2013;368(20):1935-1941.
5. Primary Care Coalition. Issue brief: collaboration between physicians and nurses works. Compare the education gaps between primary care physicians and nurse practitioners. http://www.tafp.org/Media/Default/Downloads/advocacy/scope-education.pdf. Published November 1, 2010. Accessed October 11, 2017.
6. American Medical Association. Issue brief: independent nursing practice. https://www.ama-assn.org/system/files/media-browser/premium/arc/ama-issue-brief-independent-nursing-practice.pdf. Updated 2017.
7. Tabor J, Jennings N, Kohler L, et al. The supply of physician assistants, nurse practitioners, and certified nurse midwives in Arizona. University of Arizona. http://azahec.uahs.arizona.edu/sites/default/files/u9/supply_of_pa_np_cnm.pdf. Accessed October 11, 2017.
8. Hanrahan NP, Hartley D. Employment of advanced-practice psychiatric nurses to stem rural mental health workforce shortages. Psychiatr Serv. 2008;59(1):109-111.
9. 2017 NRMP Main Residency Match the largest match on record [press release]. Washington, DC: National Resident Matching Program; March 17, 2017. http://www.nrmp.org/press-release-2017-nrmp-main-residency-match-the-largest-match-on-record. Accessed October 11, 2017.
10. Resident Physician Shortage Reduction Act of 2017, HR 2267, 115th Cong, 1st session (2017).
The author responds
I would like to thank Dr. Kendall for her passionate letter about my editorial and provide the following response. I neither asserted the equivalency of doctors and nurses or that APRNs can do what MDs do. Rather, APRNs are educated to provide highly qualified, specialty-specific advanced practice nursing, according to the tightly regulated scope of practice defined by individual states. As stated in my editorial, psychiatric mental health (PMH) APRNs engage in the practice of advanced practice PMH nursing. Is there overlap with medicine, social work, and psychology? Of course, but we are not criticized by social workers and psychologists when we engage in various psychotherapeutic approaches; rather, we are collegial and refer to each other. Why are we criticized by physicians when we prescribe from our tightly regulated legend drugs or conduct a psychiatric intake and develop a differential diagnosis and formulation that may save a life in the absence of an available psychiatrist? I would offer that PMH-APRNs are proud of their vast history of collegial relationships with psychiatrists, and that in states where turf is not an issue, there is remarkable respect and mutual referrals based on the ultimate need of finding the most appropriate care for a patient and/or family struggling to live with a psychiatric disorder.
Currently, 26 states have legislated independent practice for APRNs. This legislation was passed after decades of compiling data on the safety and efficacy of patient care outcomes in those states, and then was submitted as testimony to the legislature. State legislature decisions often are influenced by the fact that malpractice claims are decreased in areas where APRNs are independent and increased when APRNs are associated with MDs. A 2009 study1 found that between 1991 and 2007—the first 17 years that the National Practitioner Data Bank was in operation—payments were made on behalf of 37% of physicians but only 3.1% of physician assistants (PAs) and 1.5% of nurse practitioners. The study concluded: “There were no observations or trends to suggest that PAs and APNs increase liability. If anything, they may decrease the rate of reporting malpractice and adverse events.”1
To respond to Dr. Kendall’s comment, “nursing was never designed to be an autonomous practice,” nursing at the entry level of registration was originally conceived by Florence Nightingale as an autonomous profession working side-by-side with physicians, each performing different yet complementary aspects of patient care, each answering to a different hierarchy. Her work in the Crimean War attests to the positive effects of nursing on saving soldiers’ lives, which was heretofore unknown due to all the measures she initiated and meticulously documented. This autonomy, however, was gradually usurped in the private sector. Comparing RNs with MDs is like comparing apples with oranges. We would need to compare all MDs with the 3.4 million registered nurses in the United States, and that is not what my editorial addressed.
For >50 years, master’s prepared advanced practice nurses in psychiatry have been independent in their ability to have private practices, initially focusing on the provision of individual, group, and family psychotherapy. Psychiatrists did not object to this because it opened services they were unable to provide. As psychopharmacologic treatments for psychiatric disorders emerged, APRNs who had the minimum of a master’s degree and substantial psychopharmacology education, which was mandated and regulated by states, were gradually allowed to prescribe starting in the late 1970s. Most typically, these practices were in collaboration with or under supervision of an MD, but as data and outcomes were collected, legislatures began to drop this requirement.
Regarding hours, we could compare the >2,000 classroom and clinical hours and years of clinical experience accumulated by PMH-APRNs in their undergraduate and graduate psychiatric nursing curricula with the 60-hour Psychiatric Medicine course taken in the second semester of the first year of medical school.2 For many physicians, this often is the only psychiatric education they receive when going into primary care. When we consider that 70% of psychiatric care is now provided in a primary care setting, we all should be concerned and be attempting to recruit highly qualified PMH-APRNs to assist in the development and delivery of integrated primary care.
Regarding APRNs working in rural areas, Hanrahan and Hartley3 found that psychiatric APRNs were more likely than psychiatrists to live in rural areas. I contend that the issue is not the zip code of the psychiatric APRN, but rather the need to fix the problem of providers not being drawn to practice in rural and underserved populations due to salary.
Promoting autonomy for PMH-APRNs in all states is not the only way to solve the provider supply shortage, but it is a reasonable way. Unfortunately, there will be a shortage of psychiatric providers no matter what we do. Those of us who are dedicated to providing care to this vulnerable population should be finding ways to maximize our efforts and efficiencies to lessen the critical shortage. Anything less only adds to the problem and sends a negative message to the public. If we psychiatric providers cannot be supportive of each discipline practicing to the full scope and authority of their hard-earned licenses, then we are saying that we are more interested in protecting turf than providing desperately needed care.
Mary D. Moller, DNP, ARNP, PMHCNS-BC, CPRP, FAAN
Associate Professor and Coordinator PMH-DNP ProgramPacific Lutheran University School of Nursing
Director of Psychiatric Services
Northwest Integrated Health
Tacoma, Washington
References
1. Hooker RS, Nicholson JG, Le T. Does the employment of physician assistants and nurse practitioners increase liability? Journal of Medical Licensure and Discipline. 2009;95(2): 6-16.
2. Columbia University Medical Center. Medical student education in psychiatry. https://www.columbiapsychiatry.org/education-and-training/medical-student-education-psychiatry. Accessed November 16, 2017.
3. Hanrahan NP, Hartley D. Employment of advanced-practice psychiatric nurses to stem rural mental health workforce shortages. Psychiatr Serv. 2008;59(1):109-111.
In Dr. Mary Moller’s Guest Editorial “Advancing the role of advanced practice psychiatric nurses in today’s psychiatric workforce” (
Dr. Moller cited a source from the Federal Trade Commission1 that encourages the autonomous practice of APRNs to increase competition. This again implies the false equivalency between physicians and APRNs. Competition implies that the players are providing the same service. If, as nurse practitioners argue, they practice “nursing,” then they are not practicing “medicine.” Physicians and APRNs do not have the same background. Although both are charged with the care of patients, nursing is not medicine, nor should it be. Both are important and needed, but nursing was never designed to be an autonomous practice. According to the American Association of Colleges of Nursing, “Nursing and medicine are distinct health disciplines that prepare clinicians to assume different roles and meet different practice expectations.”2 In fact, the curriculum and requirements to become an APRN vary depending on the program, and some programs do not even require a BSN.3 There are online programs available for earning an APRN degree. Additionally, APRNs are only required to have 500 to 700 total hours of patient care,4 compared with the >10,000 hours physicians have once they have finished a 3-year residency, which when combined with their education amounts to >20,000 hours.5 This doesn’t account for those who have longer residencies or fellowships to further specialize in their area of training.
Dr. Moller’s main argument is that there is a dire shortage of psychiatrists and that the only way to meet this need for more providers is to make APRNs autonomous. However, no data indicate that autonomous practice of mid-level providers leads to an influx of these providers in rural areas, where the need would be greatest. Although current data on this are quite sparse, some studies indicate that the majority practice in urban areas, even in states with independent practice authority.6,7 Dr. Moller cites a source that only reviewed home zip codes of psychiatric APRNs but did not include zip codes of employment.8 Only 13% of psychiatric APRNs live in rural areas across the United States. Therefore, it is a false assertion to state that these APRNs are found primarily in rural and less populated urban areas. It is also false to imply and assume that these APRNs practice in the rural areas.
In 2017, there were 43,157 registered physician applications, with 35,969 active applications for 31,757 residency positions in the United States, and at least 11,400 medical school graduates were unmatched.9 Imagine how much more we could serve our patients by matching these graduates, whose training far surpasses that of a mid-level provider. The Resident Physician Shortage Reduction Act of 2017 aims to address this problem by increasing Medicare-funded graduate medical education (GME) residency programs in the United States.10 We can make a difference by contacting our members of Congress to encourage them to support this bill. In addition, the AMA is advocating to save funding for GME and provides an easy-to-use Web site (https://savegme.org/take-action) to contact your legislators directly to show your support for GME.
Nurse practitioners have tremendous value when their role is a part of a team; however, they should not practice without supervision, and physicians who supervise them absolutely should be providing adequate supervision. I applaud the APA and the AMA for standing up for the practice of medicine and for our patients. I hope that they continue to do so, and I encourage them to increase their efforts.
Laura Kendall, MD
Assistant Professor of Clinical Psychiatry
Department of Psychiatry and Behavioral Sciences
Keck School of Medicine
University of Southern California
Los Angeles, California
References
1. Koslov T; Office of Policy Planning. The doctor (or nurse practitioner) will see you now: competition and the regulation of advanced practice nurses. Federal Trade Commission. https://www.ftc.gov/news-events/blogs/competition-matters/2014/03/doctor-or-nurse-practitioner-will-see-you-now. Published March 7, 2014. Accessed July 26, 2017.
2. American Association of Colleges of Nursing. DNP talking points. http://www.aacnnursing.org/DNP/about/talking-points. Updated July, 2014. Accessed August 12, 2017.
3. Keyes L. MSN without a BSN? MastersInNursing.com. https://www.mastersinnursing.com/msn-without-a-bsn. Accessed August 12, 2017.
4. Iglehart JK. Expanding the role of advanced nurse practitioners—risks and rewards. New Engl J Med. 2013;368(20):1935-1941.
5. Primary Care Coalition. Issue brief: collaboration between physicians and nurses works. Compare the education gaps between primary care physicians and nurse practitioners. http://www.tafp.org/Media/Default/Downloads/advocacy/scope-education.pdf. Published November 1, 2010. Accessed October 11, 2017.
6. American Medical Association. Issue brief: independent nursing practice. https://www.ama-assn.org/system/files/media-browser/premium/arc/ama-issue-brief-independent-nursing-practice.pdf. Updated 2017.
7. Tabor J, Jennings N, Kohler L, et al. The supply of physician assistants, nurse practitioners, and certified nurse midwives in Arizona. University of Arizona. http://azahec.uahs.arizona.edu/sites/default/files/u9/supply_of_pa_np_cnm.pdf. Accessed October 11, 2017.
8. Hanrahan NP, Hartley D. Employment of advanced-practice psychiatric nurses to stem rural mental health workforce shortages. Psychiatr Serv. 2008;59(1):109-111.
9. 2017 NRMP Main Residency Match the largest match on record [press release]. Washington, DC: National Resident Matching Program; March 17, 2017. http://www.nrmp.org/press-release-2017-nrmp-main-residency-match-the-largest-match-on-record. Accessed October 11, 2017.
10. Resident Physician Shortage Reduction Act of 2017, HR 2267, 115th Cong, 1st session (2017).
The author responds
I would like to thank Dr. Kendall for her passionate letter about my editorial and provide the following response. I neither asserted the equivalency of doctors and nurses or that APRNs can do what MDs do. Rather, APRNs are educated to provide highly qualified, specialty-specific advanced practice nursing, according to the tightly regulated scope of practice defined by individual states. As stated in my editorial, psychiatric mental health (PMH) APRNs engage in the practice of advanced practice PMH nursing. Is there overlap with medicine, social work, and psychology? Of course, but we are not criticized by social workers and psychologists when we engage in various psychotherapeutic approaches; rather, we are collegial and refer to each other. Why are we criticized by physicians when we prescribe from our tightly regulated legend drugs or conduct a psychiatric intake and develop a differential diagnosis and formulation that may save a life in the absence of an available psychiatrist? I would offer that PMH-APRNs are proud of their vast history of collegial relationships with psychiatrists, and that in states where turf is not an issue, there is remarkable respect and mutual referrals based on the ultimate need of finding the most appropriate care for a patient and/or family struggling to live with a psychiatric disorder.
Currently, 26 states have legislated independent practice for APRNs. This legislation was passed after decades of compiling data on the safety and efficacy of patient care outcomes in those states, and then was submitted as testimony to the legislature. State legislature decisions often are influenced by the fact that malpractice claims are decreased in areas where APRNs are independent and increased when APRNs are associated with MDs. A 2009 study1 found that between 1991 and 2007—the first 17 years that the National Practitioner Data Bank was in operation—payments were made on behalf of 37% of physicians but only 3.1% of physician assistants (PAs) and 1.5% of nurse practitioners. The study concluded: “There were no observations or trends to suggest that PAs and APNs increase liability. If anything, they may decrease the rate of reporting malpractice and adverse events.”1
To respond to Dr. Kendall’s comment, “nursing was never designed to be an autonomous practice,” nursing at the entry level of registration was originally conceived by Florence Nightingale as an autonomous profession working side-by-side with physicians, each performing different yet complementary aspects of patient care, each answering to a different hierarchy. Her work in the Crimean War attests to the positive effects of nursing on saving soldiers’ lives, which was heretofore unknown due to all the measures she initiated and meticulously documented. This autonomy, however, was gradually usurped in the private sector. Comparing RNs with MDs is like comparing apples with oranges. We would need to compare all MDs with the 3.4 million registered nurses in the United States, and that is not what my editorial addressed.
For >50 years, master’s prepared advanced practice nurses in psychiatry have been independent in their ability to have private practices, initially focusing on the provision of individual, group, and family psychotherapy. Psychiatrists did not object to this because it opened services they were unable to provide. As psychopharmacologic treatments for psychiatric disorders emerged, APRNs who had the minimum of a master’s degree and substantial psychopharmacology education, which was mandated and regulated by states, were gradually allowed to prescribe starting in the late 1970s. Most typically, these practices were in collaboration with or under supervision of an MD, but as data and outcomes were collected, legislatures began to drop this requirement.
Regarding hours, we could compare the >2,000 classroom and clinical hours and years of clinical experience accumulated by PMH-APRNs in their undergraduate and graduate psychiatric nursing curricula with the 60-hour Psychiatric Medicine course taken in the second semester of the first year of medical school.2 For many physicians, this often is the only psychiatric education they receive when going into primary care. When we consider that 70% of psychiatric care is now provided in a primary care setting, we all should be concerned and be attempting to recruit highly qualified PMH-APRNs to assist in the development and delivery of integrated primary care.
Regarding APRNs working in rural areas, Hanrahan and Hartley3 found that psychiatric APRNs were more likely than psychiatrists to live in rural areas. I contend that the issue is not the zip code of the psychiatric APRN, but rather the need to fix the problem of providers not being drawn to practice in rural and underserved populations due to salary.
Promoting autonomy for PMH-APRNs in all states is not the only way to solve the provider supply shortage, but it is a reasonable way. Unfortunately, there will be a shortage of psychiatric providers no matter what we do. Those of us who are dedicated to providing care to this vulnerable population should be finding ways to maximize our efforts and efficiencies to lessen the critical shortage. Anything less only adds to the problem and sends a negative message to the public. If we psychiatric providers cannot be supportive of each discipline practicing to the full scope and authority of their hard-earned licenses, then we are saying that we are more interested in protecting turf than providing desperately needed care.
Mary D. Moller, DNP, ARNP, PMHCNS-BC, CPRP, FAAN
Associate Professor and Coordinator PMH-DNP ProgramPacific Lutheran University School of Nursing
Director of Psychiatric Services
Northwest Integrated Health
Tacoma, Washington
References
1. Hooker RS, Nicholson JG, Le T. Does the employment of physician assistants and nurse practitioners increase liability? Journal of Medical Licensure and Discipline. 2009;95(2): 6-16.
2. Columbia University Medical Center. Medical student education in psychiatry. https://www.columbiapsychiatry.org/education-and-training/medical-student-education-psychiatry. Accessed November 16, 2017.
3. Hanrahan NP, Hartley D. Employment of advanced-practice psychiatric nurses to stem rural mental health workforce shortages. Psychiatr Serv. 2008;59(1):109-111.
In Dr. Mary Moller’s Guest Editorial “Advancing the role of advanced practice psychiatric nurses in today’s psychiatric workforce” (
Dr. Moller cited a source from the Federal Trade Commission1 that encourages the autonomous practice of APRNs to increase competition. This again implies the false equivalency between physicians and APRNs. Competition implies that the players are providing the same service. If, as nurse practitioners argue, they practice “nursing,” then they are not practicing “medicine.” Physicians and APRNs do not have the same background. Although both are charged with the care of patients, nursing is not medicine, nor should it be. Both are important and needed, but nursing was never designed to be an autonomous practice. According to the American Association of Colleges of Nursing, “Nursing and medicine are distinct health disciplines that prepare clinicians to assume different roles and meet different practice expectations.”2 In fact, the curriculum and requirements to become an APRN vary depending on the program, and some programs do not even require a BSN.3 There are online programs available for earning an APRN degree. Additionally, APRNs are only required to have 500 to 700 total hours of patient care,4 compared with the >10,000 hours physicians have once they have finished a 3-year residency, which when combined with their education amounts to >20,000 hours.5 This doesn’t account for those who have longer residencies or fellowships to further specialize in their area of training.
Dr. Moller’s main argument is that there is a dire shortage of psychiatrists and that the only way to meet this need for more providers is to make APRNs autonomous. However, no data indicate that autonomous practice of mid-level providers leads to an influx of these providers in rural areas, where the need would be greatest. Although current data on this are quite sparse, some studies indicate that the majority practice in urban areas, even in states with independent practice authority.6,7 Dr. Moller cites a source that only reviewed home zip codes of psychiatric APRNs but did not include zip codes of employment.8 Only 13% of psychiatric APRNs live in rural areas across the United States. Therefore, it is a false assertion to state that these APRNs are found primarily in rural and less populated urban areas. It is also false to imply and assume that these APRNs practice in the rural areas.
In 2017, there were 43,157 registered physician applications, with 35,969 active applications for 31,757 residency positions in the United States, and at least 11,400 medical school graduates were unmatched.9 Imagine how much more we could serve our patients by matching these graduates, whose training far surpasses that of a mid-level provider. The Resident Physician Shortage Reduction Act of 2017 aims to address this problem by increasing Medicare-funded graduate medical education (GME) residency programs in the United States.10 We can make a difference by contacting our members of Congress to encourage them to support this bill. In addition, the AMA is advocating to save funding for GME and provides an easy-to-use Web site (https://savegme.org/take-action) to contact your legislators directly to show your support for GME.
Nurse practitioners have tremendous value when their role is a part of a team; however, they should not practice without supervision, and physicians who supervise them absolutely should be providing adequate supervision. I applaud the APA and the AMA for standing up for the practice of medicine and for our patients. I hope that they continue to do so, and I encourage them to increase their efforts.
Laura Kendall, MD
Assistant Professor of Clinical Psychiatry
Department of Psychiatry and Behavioral Sciences
Keck School of Medicine
University of Southern California
Los Angeles, California
References
1. Koslov T; Office of Policy Planning. The doctor (or nurse practitioner) will see you now: competition and the regulation of advanced practice nurses. Federal Trade Commission. https://www.ftc.gov/news-events/blogs/competition-matters/2014/03/doctor-or-nurse-practitioner-will-see-you-now. Published March 7, 2014. Accessed July 26, 2017.
2. American Association of Colleges of Nursing. DNP talking points. http://www.aacnnursing.org/DNP/about/talking-points. Updated July, 2014. Accessed August 12, 2017.
3. Keyes L. MSN without a BSN? MastersInNursing.com. https://www.mastersinnursing.com/msn-without-a-bsn. Accessed August 12, 2017.
4. Iglehart JK. Expanding the role of advanced nurse practitioners—risks and rewards. New Engl J Med. 2013;368(20):1935-1941.
5. Primary Care Coalition. Issue brief: collaboration between physicians and nurses works. Compare the education gaps between primary care physicians and nurse practitioners. http://www.tafp.org/Media/Default/Downloads/advocacy/scope-education.pdf. Published November 1, 2010. Accessed October 11, 2017.
6. American Medical Association. Issue brief: independent nursing practice. https://www.ama-assn.org/system/files/media-browser/premium/arc/ama-issue-brief-independent-nursing-practice.pdf. Updated 2017.
7. Tabor J, Jennings N, Kohler L, et al. The supply of physician assistants, nurse practitioners, and certified nurse midwives in Arizona. University of Arizona. http://azahec.uahs.arizona.edu/sites/default/files/u9/supply_of_pa_np_cnm.pdf. Accessed October 11, 2017.
8. Hanrahan NP, Hartley D. Employment of advanced-practice psychiatric nurses to stem rural mental health workforce shortages. Psychiatr Serv. 2008;59(1):109-111.
9. 2017 NRMP Main Residency Match the largest match on record [press release]. Washington, DC: National Resident Matching Program; March 17, 2017. http://www.nrmp.org/press-release-2017-nrmp-main-residency-match-the-largest-match-on-record. Accessed October 11, 2017.
10. Resident Physician Shortage Reduction Act of 2017, HR 2267, 115th Cong, 1st session (2017).
The author responds
I would like to thank Dr. Kendall for her passionate letter about my editorial and provide the following response. I neither asserted the equivalency of doctors and nurses or that APRNs can do what MDs do. Rather, APRNs are educated to provide highly qualified, specialty-specific advanced practice nursing, according to the tightly regulated scope of practice defined by individual states. As stated in my editorial, psychiatric mental health (PMH) APRNs engage in the practice of advanced practice PMH nursing. Is there overlap with medicine, social work, and psychology? Of course, but we are not criticized by social workers and psychologists when we engage in various psychotherapeutic approaches; rather, we are collegial and refer to each other. Why are we criticized by physicians when we prescribe from our tightly regulated legend drugs or conduct a psychiatric intake and develop a differential diagnosis and formulation that may save a life in the absence of an available psychiatrist? I would offer that PMH-APRNs are proud of their vast history of collegial relationships with psychiatrists, and that in states where turf is not an issue, there is remarkable respect and mutual referrals based on the ultimate need of finding the most appropriate care for a patient and/or family struggling to live with a psychiatric disorder.
Currently, 26 states have legislated independent practice for APRNs. This legislation was passed after decades of compiling data on the safety and efficacy of patient care outcomes in those states, and then was submitted as testimony to the legislature. State legislature decisions often are influenced by the fact that malpractice claims are decreased in areas where APRNs are independent and increased when APRNs are associated with MDs. A 2009 study1 found that between 1991 and 2007—the first 17 years that the National Practitioner Data Bank was in operation—payments were made on behalf of 37% of physicians but only 3.1% of physician assistants (PAs) and 1.5% of nurse practitioners. The study concluded: “There were no observations or trends to suggest that PAs and APNs increase liability. If anything, they may decrease the rate of reporting malpractice and adverse events.”1
To respond to Dr. Kendall’s comment, “nursing was never designed to be an autonomous practice,” nursing at the entry level of registration was originally conceived by Florence Nightingale as an autonomous profession working side-by-side with physicians, each performing different yet complementary aspects of patient care, each answering to a different hierarchy. Her work in the Crimean War attests to the positive effects of nursing on saving soldiers’ lives, which was heretofore unknown due to all the measures she initiated and meticulously documented. This autonomy, however, was gradually usurped in the private sector. Comparing RNs with MDs is like comparing apples with oranges. We would need to compare all MDs with the 3.4 million registered nurses in the United States, and that is not what my editorial addressed.
For >50 years, master’s prepared advanced practice nurses in psychiatry have been independent in their ability to have private practices, initially focusing on the provision of individual, group, and family psychotherapy. Psychiatrists did not object to this because it opened services they were unable to provide. As psychopharmacologic treatments for psychiatric disorders emerged, APRNs who had the minimum of a master’s degree and substantial psychopharmacology education, which was mandated and regulated by states, were gradually allowed to prescribe starting in the late 1970s. Most typically, these practices were in collaboration with or under supervision of an MD, but as data and outcomes were collected, legislatures began to drop this requirement.
Regarding hours, we could compare the >2,000 classroom and clinical hours and years of clinical experience accumulated by PMH-APRNs in their undergraduate and graduate psychiatric nursing curricula with the 60-hour Psychiatric Medicine course taken in the second semester of the first year of medical school.2 For many physicians, this often is the only psychiatric education they receive when going into primary care. When we consider that 70% of psychiatric care is now provided in a primary care setting, we all should be concerned and be attempting to recruit highly qualified PMH-APRNs to assist in the development and delivery of integrated primary care.
Regarding APRNs working in rural areas, Hanrahan and Hartley3 found that psychiatric APRNs were more likely than psychiatrists to live in rural areas. I contend that the issue is not the zip code of the psychiatric APRN, but rather the need to fix the problem of providers not being drawn to practice in rural and underserved populations due to salary.
Promoting autonomy for PMH-APRNs in all states is not the only way to solve the provider supply shortage, but it is a reasonable way. Unfortunately, there will be a shortage of psychiatric providers no matter what we do. Those of us who are dedicated to providing care to this vulnerable population should be finding ways to maximize our efforts and efficiencies to lessen the critical shortage. Anything less only adds to the problem and sends a negative message to the public. If we psychiatric providers cannot be supportive of each discipline practicing to the full scope and authority of their hard-earned licenses, then we are saying that we are more interested in protecting turf than providing desperately needed care.
Mary D. Moller, DNP, ARNP, PMHCNS-BC, CPRP, FAAN
Associate Professor and Coordinator PMH-DNP ProgramPacific Lutheran University School of Nursing
Director of Psychiatric Services
Northwest Integrated Health
Tacoma, Washington
References
1. Hooker RS, Nicholson JG, Le T. Does the employment of physician assistants and nurse practitioners increase liability? Journal of Medical Licensure and Discipline. 2009;95(2): 6-16.
2. Columbia University Medical Center. Medical student education in psychiatry. https://www.columbiapsychiatry.org/education-and-training/medical-student-education-psychiatry. Accessed November 16, 2017.
3. Hanrahan NP, Hartley D. Employment of advanced-practice psychiatric nurses to stem rural mental health workforce shortages. Psychiatr Serv. 2008;59(1):109-111.
Time for another cup of coffee … or two
Nutrition enthusiasts have been saying for years that “we are what we eat.” In this issue of JFP, Drs. Malone and Tsai review the evidence regarding the medicinal properties of certain herbal and botanical supplements. (See here.) Although there are many unfounded claims about the health effects of a wide variety of supplements, increasing evidence from well-conducted randomized trials and large epidemiologic studies demonstrates that certain items we ingest have therapeutic value for a variety of conditions.
However, as Dr. Malone points out, herbal supplements and botanicals are not regulated by the US Food and Drug Administration, so manufacturers are not required to provide proof of effectiveness or safety to market their products. Some of these products have adverse effects. For example, butterbur can cause liver toxicity.1
Because about 38% of Americans use supplements, all primary care clinicians should know which products do—and do not—have evidence of efficacy.2 I suggest you read Dr. Malone’s 2-part article (part 2 is available here), but I can’t resist highlighting several of my favorites:
Coffee. Coffee used to be maligned because of its caffeine content, but more recent evidence suggests it protects against liver disease and has positive effects on cardiovascular disease and even mortality. (See Dr. Malone’s article for references.) There are no randomized trials, however, so we can’t be entirely sure if these associations are causal, but at least I can feel better about the 4 cups of coffee I drink every day!
Tea, especially green tea, appears to have many positive effects on health, including potential roles in reducing the risk of cancer, cardiovascular disease, type 2 diabetes, and even dementia. As with coffee, these associations are based on large observational studies and not randomized trials.
Chamomile. If your coffee gives you too much of a buzz and causes you to feel anxious, calm down with chamomile tea or oil. Evidence from randomized trials indicates it has positive effects on insomnia and anxiety.
Peppermint oil. The data for calming irritable bowel syndrome is fairly strong, and it may be effective in aborting migraines when applied to the forehead. It’s certainly worth a try for these difficult-to-treat conditions.
When patients ask you about botanicals and herbals, a great resource is
Now I will finish my fourth cup of coffee . . .
1. Daniel O, Mauskop A. Nutraceuticals in acute and prophylactic treatment of migraine. Curr Treat Options Neurol. 2016;18:14.
2. National Center for Complementary and Integrative Health. The use of complementary and alternative medicine in the United States. Available at: https://nccih.nih.gov/research/statistics/2007/camsurvey_fs1.htm. Accessed December 14, 2017.
Nutrition enthusiasts have been saying for years that “we are what we eat.” In this issue of JFP, Drs. Malone and Tsai review the evidence regarding the medicinal properties of certain herbal and botanical supplements. (See here.) Although there are many unfounded claims about the health effects of a wide variety of supplements, increasing evidence from well-conducted randomized trials and large epidemiologic studies demonstrates that certain items we ingest have therapeutic value for a variety of conditions.
However, as Dr. Malone points out, herbal supplements and botanicals are not regulated by the US Food and Drug Administration, so manufacturers are not required to provide proof of effectiveness or safety to market their products. Some of these products have adverse effects. For example, butterbur can cause liver toxicity.1
Because about 38% of Americans use supplements, all primary care clinicians should know which products do—and do not—have evidence of efficacy.2 I suggest you read Dr. Malone’s 2-part article (part 2 is available here), but I can’t resist highlighting several of my favorites:
Coffee. Coffee used to be maligned because of its caffeine content, but more recent evidence suggests it protects against liver disease and has positive effects on cardiovascular disease and even mortality. (See Dr. Malone’s article for references.) There are no randomized trials, however, so we can’t be entirely sure if these associations are causal, but at least I can feel better about the 4 cups of coffee I drink every day!
Tea, especially green tea, appears to have many positive effects on health, including potential roles in reducing the risk of cancer, cardiovascular disease, type 2 diabetes, and even dementia. As with coffee, these associations are based on large observational studies and not randomized trials.
Chamomile. If your coffee gives you too much of a buzz and causes you to feel anxious, calm down with chamomile tea or oil. Evidence from randomized trials indicates it has positive effects on insomnia and anxiety.
Peppermint oil. The data for calming irritable bowel syndrome is fairly strong, and it may be effective in aborting migraines when applied to the forehead. It’s certainly worth a try for these difficult-to-treat conditions.
When patients ask you about botanicals and herbals, a great resource is
Now I will finish my fourth cup of coffee . . .
Nutrition enthusiasts have been saying for years that “we are what we eat.” In this issue of JFP, Drs. Malone and Tsai review the evidence regarding the medicinal properties of certain herbal and botanical supplements. (See here.) Although there are many unfounded claims about the health effects of a wide variety of supplements, increasing evidence from well-conducted randomized trials and large epidemiologic studies demonstrates that certain items we ingest have therapeutic value for a variety of conditions.
However, as Dr. Malone points out, herbal supplements and botanicals are not regulated by the US Food and Drug Administration, so manufacturers are not required to provide proof of effectiveness or safety to market their products. Some of these products have adverse effects. For example, butterbur can cause liver toxicity.1
Because about 38% of Americans use supplements, all primary care clinicians should know which products do—and do not—have evidence of efficacy.2 I suggest you read Dr. Malone’s 2-part article (part 2 is available here), but I can’t resist highlighting several of my favorites:
Coffee. Coffee used to be maligned because of its caffeine content, but more recent evidence suggests it protects against liver disease and has positive effects on cardiovascular disease and even mortality. (See Dr. Malone’s article for references.) There are no randomized trials, however, so we can’t be entirely sure if these associations are causal, but at least I can feel better about the 4 cups of coffee I drink every day!
Tea, especially green tea, appears to have many positive effects on health, including potential roles in reducing the risk of cancer, cardiovascular disease, type 2 diabetes, and even dementia. As with coffee, these associations are based on large observational studies and not randomized trials.
Chamomile. If your coffee gives you too much of a buzz and causes you to feel anxious, calm down with chamomile tea or oil. Evidence from randomized trials indicates it has positive effects on insomnia and anxiety.
Peppermint oil. The data for calming irritable bowel syndrome is fairly strong, and it may be effective in aborting migraines when applied to the forehead. It’s certainly worth a try for these difficult-to-treat conditions.
When patients ask you about botanicals and herbals, a great resource is
Now I will finish my fourth cup of coffee . . .
1. Daniel O, Mauskop A. Nutraceuticals in acute and prophylactic treatment of migraine. Curr Treat Options Neurol. 2016;18:14.
2. National Center for Complementary and Integrative Health. The use of complementary and alternative medicine in the United States. Available at: https://nccih.nih.gov/research/statistics/2007/camsurvey_fs1.htm. Accessed December 14, 2017.
1. Daniel O, Mauskop A. Nutraceuticals in acute and prophylactic treatment of migraine. Curr Treat Options Neurol. 2016;18:14.
2. National Center for Complementary and Integrative Health. The use of complementary and alternative medicine in the United States. Available at: https://nccih.nih.gov/research/statistics/2007/camsurvey_fs1.htm. Accessed December 14, 2017.
Is this hypertension treatment strategy based on SOR “A” evidence?
In the article, “Hypertension treatment strategies for older adults” (J Fam Pract. 2017;66:546-554), Hansell et al gave an “A” Strength of Recommendation (SOR) rating to the Practice Recommendation that read: “Target a systolic blood pressure (SBP) <120 mm Hg in community-dwelling, nondiabetic patients ≥75 years of age if it is achievable without undue burden.”
As justification for this SBP target, the authors cited a subgroup analysis from the Systolic Blood Pressure Intervention Trial (SPRINT),1 which consisted of patients ≥75 years of age.2 I posit that the inconsistencies of the data cited by Hansell et al contradict an “A” rating, and that the methodology used in SPRINT greatly mitigates the generalizability of the results.
First, Hansell et al admit that no consensus exists on an optimal BP target for older patients. SOR taxonomy requires that the evidence behind an SOR of “A” be based on consistent and good-quality patient-oriented evidence.3 One source cited by the authors states that evidence supporting lower targets is inconsistent,4 while a recent Cochrane review does not support low BP targets.5 Given that the evidence is inconsistent, the SOR should be a “B”, at best.
Second, the evidence to target a systolic BP <120 mm Hg primarily comes from SPRINT.1,2,4 In a Letter to the Editor that appeared in The New England Journal of Medicine, Dr. Marc A. Pfeffer addressed a key methodology issue of that trial: SPRINT protocol called for the withdrawal of antihypertensive therapy in the standard treatment group if a single systolic BP reading was <130 mm Hg, or if readings at 2 or more consecutive visits were <135 mm Hg, regardless of patient symptoms.6
The letter also questioned how frequently this withdrawal occurred, to which the SPRINT authors replied that 87% of participants required at least one reduction in the dose of medication to maintain the treatment target in the standard group, and complete withdrawal of medication was required in <7.5% of participants.7 While this dose adjustment may have been necessary to adequately test the SPRINT hypothesis that lower systolic BP targets are better, routine dose reduction in an asymptomatic patient is not standard practice.
Given the small benefit in absolute risk reduction in SPRINT’s aggressive hypertensive treatment arm of 0.54% per year for the primary composite outcome and 0.37% per year for all-cause mortality,2 the frequent medication dose reductions in the standard treatment arm likely contributed significantly to the statistical benefit seen in the aggressive treatment group in SPRINT.
If an SOR of “A” for BP targets is to be made, the print publication of Hansell et al’s article should communicate the degree of benefit, preferably in terms of absolute risk reduction. Only the online publication of TABLE W1 stated the degree of benefit in the SPRINT subgroup study, but it was stated in terms of relative risk.
Given the current suboptimal rates of hypertension control, primary care physicians would do well to impact morbidity and mortality in older adults by working to achieve standard targets, such as an SBP of <140 mm Hg or <150 mm Hg. Once standard targets are achieved, a conversation could then ensue about the potential benefits and harms of lower BP targets.
Chris Fallert, MD
St. Paul, Minn
1. Williamson JD, Supiano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
2. SPRINT Research Group, Wright JT Jr, Williamson JD, et al. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2015;373:2103-2106.
3. Ebell MH, Siwek J, Weiss BD, et al. Strength of recommendation taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Physician. 2004;69:548-556.
4. Weiss J, Freeman M, Low A, et al. Benefits and harms of intensive blood pressure treatment in adults aged 60 years or older: a systematic review and meta-analysis. Ann Intern Med. 2017;166:419-429.
5. Garrison SR, Kolber MR, Korownyk CS, et al. Blood pressure targets for hypertension in older adults. Cochrane Database Syst Rev. 2017;8:CD011575.
6. Pfeffer MA. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2016;374:2290.
7. Wright JT Jr, Whelton PK, Reboussin DM. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2016;374:2294.
Authors’ response:
An SOR of “A” is based on consistent and good-quality patient-oriented evidence, which is further defined for treatment, prevention, and screening studies as (a) systematic reviews/meta-analyses of randomized controlled trials (RCTs) with consistent findings or (b) a high-quality individual RCT.1 The recommendation to “target a systolic blood pressure (BP) <120 mm Hg in community-dwelling, nondiabetic patients ≥75 years of age if it is achievable without undue burden” meets level 1 evidence based on both (a) and (b).
While a Cochrane review of hypertension did not support a systolic BP target <120 mm Hg, the populations evaluated included a variety of ages; the studies did not specifically focus on those ≥75 years of age with inherently high cardiovascular risk while excluding patients with diabetes.2 The Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial results,3 which are often viewed as inconsistent with SPRINT,4 included patients with diabetes and patients of a younger average age than SPRINT. Although no overall mortality benefit of intensive BP control was found in the ACCORD trial, there was significant reduction in stroke, as well as additional benefit in the ACCORD standard glycemia group.3,5
The American College of Cardiology/American Heart Association 2017 BP guidelines summarize several meta-analyses that consistently support tighter BP control with recommendations for a lower BP target of <130 mm Hg systolic.5,6 They selected a target of <130 mm Hg, rather than <120 mm Hg, assuming that general health care providers cannot be as efficacious at lowering BP as researchers in efficacy trials.5
With regard to medication withdrawal as a flaw in the SPRINT design,4 an accepted geriatric principle is reduction in polypharmacy whenever possible. Medication reduction or withdrawal when a patient is too far below target is prudent. The 2 different target groups in an RCT have to be statistically different to draw conclusions about the differences. This strategy has been employed in other BP trials. Medication withdrawal is an appropriate means to achieve targets, which the SPRINT investigators did successfully with a least-square mean systolic BP for patients ≥75 years of age in the control group of 134.8 mm Hg and 123.4 mm Hg in the intensive group.4 Even with reduction in polypharmacy in the standard group, SPRINT demonstrated cardiovascular and mortality benefit with tighter control.4
With regard to Dr. Fallert’s comments about a small absolute risk reduction for the entire SPRINT study population, our article in JFP specifically pertains to adults ≥75 years of age. The numbers needed to treat for composite cardiovascular outcomes and all-cause mortality in the ≥75 SPRINT group are 27 (95% confidence interval [CI], 19-61) and 41 (95% CI, 27-145), respectively.4
We agree that there is suboptimal hypertension control at present. However, physicians should not only focus on this modifiable risk factor to reduce CVD risk and mortality in appropriate patients, but they should focus on doing it to the evidence-based goal.
Maggie W. Hansell, MD; Emily M. Mann, MD; Julienne K. Kirk, PharmD
Winston-Salem, NC
1. Ebell MH, Siwek J, Weiss BD, et al. Strength of recommendation taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Physician. 2004;69:548-556.
2. Garrison SR, Kolber MR, Korownyk CS, et al. Blood pressure targets for hypertension in older adults. Cochrane Database Syst Rev. 2017;8:CD011575.
3. The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive blood-pressure control in type 2 diabetes. N Engl J Med. 2010;362:1575-1585.
4. Williamson JD, Suplano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
5. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Available at: http://hyper.ahajournals.org/content/hypertensionaha/early/2017/11/10/HYP.0000000000000066.full.pdf. Accessed December 12, 2017.
6. Reboussin DM, Allen NB, Griswold ME, et al. Systematic Review for the 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the prevention, detection, evaluation, and management of high blood pressure in adults. Available at: http://hyper.ahajournals.org/content/early/2017/11/10/HYP.0000000000000067. Accessed December 12, 2017.
In the article, “Hypertension treatment strategies for older adults” (J Fam Pract. 2017;66:546-554), Hansell et al gave an “A” Strength of Recommendation (SOR) rating to the Practice Recommendation that read: “Target a systolic blood pressure (SBP) <120 mm Hg in community-dwelling, nondiabetic patients ≥75 years of age if it is achievable without undue burden.”
As justification for this SBP target, the authors cited a subgroup analysis from the Systolic Blood Pressure Intervention Trial (SPRINT),1 which consisted of patients ≥75 years of age.2 I posit that the inconsistencies of the data cited by Hansell et al contradict an “A” rating, and that the methodology used in SPRINT greatly mitigates the generalizability of the results.
First, Hansell et al admit that no consensus exists on an optimal BP target for older patients. SOR taxonomy requires that the evidence behind an SOR of “A” be based on consistent and good-quality patient-oriented evidence.3 One source cited by the authors states that evidence supporting lower targets is inconsistent,4 while a recent Cochrane review does not support low BP targets.5 Given that the evidence is inconsistent, the SOR should be a “B”, at best.
Second, the evidence to target a systolic BP <120 mm Hg primarily comes from SPRINT.1,2,4 In a Letter to the Editor that appeared in The New England Journal of Medicine, Dr. Marc A. Pfeffer addressed a key methodology issue of that trial: SPRINT protocol called for the withdrawal of antihypertensive therapy in the standard treatment group if a single systolic BP reading was <130 mm Hg, or if readings at 2 or more consecutive visits were <135 mm Hg, regardless of patient symptoms.6
The letter also questioned how frequently this withdrawal occurred, to which the SPRINT authors replied that 87% of participants required at least one reduction in the dose of medication to maintain the treatment target in the standard group, and complete withdrawal of medication was required in <7.5% of participants.7 While this dose adjustment may have been necessary to adequately test the SPRINT hypothesis that lower systolic BP targets are better, routine dose reduction in an asymptomatic patient is not standard practice.
Given the small benefit in absolute risk reduction in SPRINT’s aggressive hypertensive treatment arm of 0.54% per year for the primary composite outcome and 0.37% per year for all-cause mortality,2 the frequent medication dose reductions in the standard treatment arm likely contributed significantly to the statistical benefit seen in the aggressive treatment group in SPRINT.
If an SOR of “A” for BP targets is to be made, the print publication of Hansell et al’s article should communicate the degree of benefit, preferably in terms of absolute risk reduction. Only the online publication of TABLE W1 stated the degree of benefit in the SPRINT subgroup study, but it was stated in terms of relative risk.
Given the current suboptimal rates of hypertension control, primary care physicians would do well to impact morbidity and mortality in older adults by working to achieve standard targets, such as an SBP of <140 mm Hg or <150 mm Hg. Once standard targets are achieved, a conversation could then ensue about the potential benefits and harms of lower BP targets.
Chris Fallert, MD
St. Paul, Minn
1. Williamson JD, Supiano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
2. SPRINT Research Group, Wright JT Jr, Williamson JD, et al. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2015;373:2103-2106.
3. Ebell MH, Siwek J, Weiss BD, et al. Strength of recommendation taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Physician. 2004;69:548-556.
4. Weiss J, Freeman M, Low A, et al. Benefits and harms of intensive blood pressure treatment in adults aged 60 years or older: a systematic review and meta-analysis. Ann Intern Med. 2017;166:419-429.
5. Garrison SR, Kolber MR, Korownyk CS, et al. Blood pressure targets for hypertension in older adults. Cochrane Database Syst Rev. 2017;8:CD011575.
6. Pfeffer MA. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2016;374:2290.
7. Wright JT Jr, Whelton PK, Reboussin DM. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2016;374:2294.
Authors’ response:
An SOR of “A” is based on consistent and good-quality patient-oriented evidence, which is further defined for treatment, prevention, and screening studies as (a) systematic reviews/meta-analyses of randomized controlled trials (RCTs) with consistent findings or (b) a high-quality individual RCT.1 The recommendation to “target a systolic blood pressure (BP) <120 mm Hg in community-dwelling, nondiabetic patients ≥75 years of age if it is achievable without undue burden” meets level 1 evidence based on both (a) and (b).
While a Cochrane review of hypertension did not support a systolic BP target <120 mm Hg, the populations evaluated included a variety of ages; the studies did not specifically focus on those ≥75 years of age with inherently high cardiovascular risk while excluding patients with diabetes.2 The Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial results,3 which are often viewed as inconsistent with SPRINT,4 included patients with diabetes and patients of a younger average age than SPRINT. Although no overall mortality benefit of intensive BP control was found in the ACCORD trial, there was significant reduction in stroke, as well as additional benefit in the ACCORD standard glycemia group.3,5
The American College of Cardiology/American Heart Association 2017 BP guidelines summarize several meta-analyses that consistently support tighter BP control with recommendations for a lower BP target of <130 mm Hg systolic.5,6 They selected a target of <130 mm Hg, rather than <120 mm Hg, assuming that general health care providers cannot be as efficacious at lowering BP as researchers in efficacy trials.5
With regard to medication withdrawal as a flaw in the SPRINT design,4 an accepted geriatric principle is reduction in polypharmacy whenever possible. Medication reduction or withdrawal when a patient is too far below target is prudent. The 2 different target groups in an RCT have to be statistically different to draw conclusions about the differences. This strategy has been employed in other BP trials. Medication withdrawal is an appropriate means to achieve targets, which the SPRINT investigators did successfully with a least-square mean systolic BP for patients ≥75 years of age in the control group of 134.8 mm Hg and 123.4 mm Hg in the intensive group.4 Even with reduction in polypharmacy in the standard group, SPRINT demonstrated cardiovascular and mortality benefit with tighter control.4
With regard to Dr. Fallert’s comments about a small absolute risk reduction for the entire SPRINT study population, our article in JFP specifically pertains to adults ≥75 years of age. The numbers needed to treat for composite cardiovascular outcomes and all-cause mortality in the ≥75 SPRINT group are 27 (95% confidence interval [CI], 19-61) and 41 (95% CI, 27-145), respectively.4
We agree that there is suboptimal hypertension control at present. However, physicians should not only focus on this modifiable risk factor to reduce CVD risk and mortality in appropriate patients, but they should focus on doing it to the evidence-based goal.
Maggie W. Hansell, MD; Emily M. Mann, MD; Julienne K. Kirk, PharmD
Winston-Salem, NC
1. Ebell MH, Siwek J, Weiss BD, et al. Strength of recommendation taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Physician. 2004;69:548-556.
2. Garrison SR, Kolber MR, Korownyk CS, et al. Blood pressure targets for hypertension in older adults. Cochrane Database Syst Rev. 2017;8:CD011575.
3. The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive blood-pressure control in type 2 diabetes. N Engl J Med. 2010;362:1575-1585.
4. Williamson JD, Suplano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
5. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Available at: http://hyper.ahajournals.org/content/hypertensionaha/early/2017/11/10/HYP.0000000000000066.full.pdf. Accessed December 12, 2017.
6. Reboussin DM, Allen NB, Griswold ME, et al. Systematic Review for the 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the prevention, detection, evaluation, and management of high blood pressure in adults. Available at: http://hyper.ahajournals.org/content/early/2017/11/10/HYP.0000000000000067. Accessed December 12, 2017.
In the article, “Hypertension treatment strategies for older adults” (J Fam Pract. 2017;66:546-554), Hansell et al gave an “A” Strength of Recommendation (SOR) rating to the Practice Recommendation that read: “Target a systolic blood pressure (SBP) <120 mm Hg in community-dwelling, nondiabetic patients ≥75 years of age if it is achievable without undue burden.”
As justification for this SBP target, the authors cited a subgroup analysis from the Systolic Blood Pressure Intervention Trial (SPRINT),1 which consisted of patients ≥75 years of age.2 I posit that the inconsistencies of the data cited by Hansell et al contradict an “A” rating, and that the methodology used in SPRINT greatly mitigates the generalizability of the results.
First, Hansell et al admit that no consensus exists on an optimal BP target for older patients. SOR taxonomy requires that the evidence behind an SOR of “A” be based on consistent and good-quality patient-oriented evidence.3 One source cited by the authors states that evidence supporting lower targets is inconsistent,4 while a recent Cochrane review does not support low BP targets.5 Given that the evidence is inconsistent, the SOR should be a “B”, at best.
Second, the evidence to target a systolic BP <120 mm Hg primarily comes from SPRINT.1,2,4 In a Letter to the Editor that appeared in The New England Journal of Medicine, Dr. Marc A. Pfeffer addressed a key methodology issue of that trial: SPRINT protocol called for the withdrawal of antihypertensive therapy in the standard treatment group if a single systolic BP reading was <130 mm Hg, or if readings at 2 or more consecutive visits were <135 mm Hg, regardless of patient symptoms.6
The letter also questioned how frequently this withdrawal occurred, to which the SPRINT authors replied that 87% of participants required at least one reduction in the dose of medication to maintain the treatment target in the standard group, and complete withdrawal of medication was required in <7.5% of participants.7 While this dose adjustment may have been necessary to adequately test the SPRINT hypothesis that lower systolic BP targets are better, routine dose reduction in an asymptomatic patient is not standard practice.
Given the small benefit in absolute risk reduction in SPRINT’s aggressive hypertensive treatment arm of 0.54% per year for the primary composite outcome and 0.37% per year for all-cause mortality,2 the frequent medication dose reductions in the standard treatment arm likely contributed significantly to the statistical benefit seen in the aggressive treatment group in SPRINT.
If an SOR of “A” for BP targets is to be made, the print publication of Hansell et al’s article should communicate the degree of benefit, preferably in terms of absolute risk reduction. Only the online publication of TABLE W1 stated the degree of benefit in the SPRINT subgroup study, but it was stated in terms of relative risk.
Given the current suboptimal rates of hypertension control, primary care physicians would do well to impact morbidity and mortality in older adults by working to achieve standard targets, such as an SBP of <140 mm Hg or <150 mm Hg. Once standard targets are achieved, a conversation could then ensue about the potential benefits and harms of lower BP targets.
Chris Fallert, MD
St. Paul, Minn
1. Williamson JD, Supiano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
2. SPRINT Research Group, Wright JT Jr, Williamson JD, et al. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2015;373:2103-2106.
3. Ebell MH, Siwek J, Weiss BD, et al. Strength of recommendation taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Physician. 2004;69:548-556.
4. Weiss J, Freeman M, Low A, et al. Benefits and harms of intensive blood pressure treatment in adults aged 60 years or older: a systematic review and meta-analysis. Ann Intern Med. 2017;166:419-429.
5. Garrison SR, Kolber MR, Korownyk CS, et al. Blood pressure targets for hypertension in older adults. Cochrane Database Syst Rev. 2017;8:CD011575.
6. Pfeffer MA. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2016;374:2290.
7. Wright JT Jr, Whelton PK, Reboussin DM. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2016;374:2294.
Authors’ response:
An SOR of “A” is based on consistent and good-quality patient-oriented evidence, which is further defined for treatment, prevention, and screening studies as (a) systematic reviews/meta-analyses of randomized controlled trials (RCTs) with consistent findings or (b) a high-quality individual RCT.1 The recommendation to “target a systolic blood pressure (BP) <120 mm Hg in community-dwelling, nondiabetic patients ≥75 years of age if it is achievable without undue burden” meets level 1 evidence based on both (a) and (b).
While a Cochrane review of hypertension did not support a systolic BP target <120 mm Hg, the populations evaluated included a variety of ages; the studies did not specifically focus on those ≥75 years of age with inherently high cardiovascular risk while excluding patients with diabetes.2 The Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial results,3 which are often viewed as inconsistent with SPRINT,4 included patients with diabetes and patients of a younger average age than SPRINT. Although no overall mortality benefit of intensive BP control was found in the ACCORD trial, there was significant reduction in stroke, as well as additional benefit in the ACCORD standard glycemia group.3,5
The American College of Cardiology/American Heart Association 2017 BP guidelines summarize several meta-analyses that consistently support tighter BP control with recommendations for a lower BP target of <130 mm Hg systolic.5,6 They selected a target of <130 mm Hg, rather than <120 mm Hg, assuming that general health care providers cannot be as efficacious at lowering BP as researchers in efficacy trials.5
With regard to medication withdrawal as a flaw in the SPRINT design,4 an accepted geriatric principle is reduction in polypharmacy whenever possible. Medication reduction or withdrawal when a patient is too far below target is prudent. The 2 different target groups in an RCT have to be statistically different to draw conclusions about the differences. This strategy has been employed in other BP trials. Medication withdrawal is an appropriate means to achieve targets, which the SPRINT investigators did successfully with a least-square mean systolic BP for patients ≥75 years of age in the control group of 134.8 mm Hg and 123.4 mm Hg in the intensive group.4 Even with reduction in polypharmacy in the standard group, SPRINT demonstrated cardiovascular and mortality benefit with tighter control.4
With regard to Dr. Fallert’s comments about a small absolute risk reduction for the entire SPRINT study population, our article in JFP specifically pertains to adults ≥75 years of age. The numbers needed to treat for composite cardiovascular outcomes and all-cause mortality in the ≥75 SPRINT group are 27 (95% confidence interval [CI], 19-61) and 41 (95% CI, 27-145), respectively.4
We agree that there is suboptimal hypertension control at present. However, physicians should not only focus on this modifiable risk factor to reduce CVD risk and mortality in appropriate patients, but they should focus on doing it to the evidence-based goal.
Maggie W. Hansell, MD; Emily M. Mann, MD; Julienne K. Kirk, PharmD
Winston-Salem, NC
1. Ebell MH, Siwek J, Weiss BD, et al. Strength of recommendation taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Physician. 2004;69:548-556.
2. Garrison SR, Kolber MR, Korownyk CS, et al. Blood pressure targets for hypertension in older adults. Cochrane Database Syst Rev. 2017;8:CD011575.
3. The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive blood-pressure control in type 2 diabetes. N Engl J Med. 2010;362:1575-1585.
4. Williamson JD, Suplano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
5. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Available at: http://hyper.ahajournals.org/content/hypertensionaha/early/2017/11/10/HYP.0000000000000066.full.pdf. Accessed December 12, 2017.
6. Reboussin DM, Allen NB, Griswold ME, et al. Systematic Review for the 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the prevention, detection, evaluation, and management of high blood pressure in adults. Available at: http://hyper.ahajournals.org/content/early/2017/11/10/HYP.0000000000000067. Accessed December 12, 2017.
Let’s recognize Dr. John Geyman, too
Thank you for your tribute to David Warfield Stires, The Journal of Family Practice’s founding publisher (J Fam Pract. 2017;66:654-655). The real hero of the story, however, is Dr. John Geyman, who had the vision to found a research journal at the birth of our specialty. This was no easy task, as John faced a challenging chicken-and-egg problem: how to establish a research journal when academic family medicine was just getting underway and had no track record of generating a steady stream of quality research. The latter problem was due, in part, to the lack of a research journal devoted to promoting and publishing research in the field.
Yet, John did it, putting family medicine research on the publishing map. His groundbreaking work set the stage for future journals, including the Journal of the American Board of Family Medicine, the American Medical Association’s now-defunct Archives of Family Medicine, and the American Academy of Family Physicians’ Annals of Family Medicine.
As a family medicine resident in the 1970s, I remember coveting JFP so much that I managed to collect every issue from Volume 1, Issue 1, through the turn of the century. And as a young faculty member at Georgetown University Medical Center in the 1980s, I painstakingly created an annotated bibliography of the then-published content of JFP to use for teaching, research, and administration.
When I became editor of American Family Physician in 1988, I made a pilgrimage to the University of Washington School of Medicine in Seattle, where John was chairman of the Department of Family Medicine. I wanted to seek his advice, learn from his vast experience, and pay tribute to all that he’d done for our specialty. Over the past 30 years, John has continued to leave his mark. (See http://www.johngeymanmd.org/bio.html.)
A tribute to David Warfield Stires is incomplete without a corresponding acknowledgement and celebration of John’s decades-long visionary leadership in family medicine.
Jay Siwek, MD
Washington, DC
Thank you for your tribute to David Warfield Stires, The Journal of Family Practice’s founding publisher (J Fam Pract. 2017;66:654-655). The real hero of the story, however, is Dr. John Geyman, who had the vision to found a research journal at the birth of our specialty. This was no easy task, as John faced a challenging chicken-and-egg problem: how to establish a research journal when academic family medicine was just getting underway and had no track record of generating a steady stream of quality research. The latter problem was due, in part, to the lack of a research journal devoted to promoting and publishing research in the field.
Yet, John did it, putting family medicine research on the publishing map. His groundbreaking work set the stage for future journals, including the Journal of the American Board of Family Medicine, the American Medical Association’s now-defunct Archives of Family Medicine, and the American Academy of Family Physicians’ Annals of Family Medicine.
As a family medicine resident in the 1970s, I remember coveting JFP so much that I managed to collect every issue from Volume 1, Issue 1, through the turn of the century. And as a young faculty member at Georgetown University Medical Center in the 1980s, I painstakingly created an annotated bibliography of the then-published content of JFP to use for teaching, research, and administration.
When I became editor of American Family Physician in 1988, I made a pilgrimage to the University of Washington School of Medicine in Seattle, where John was chairman of the Department of Family Medicine. I wanted to seek his advice, learn from his vast experience, and pay tribute to all that he’d done for our specialty. Over the past 30 years, John has continued to leave his mark. (See http://www.johngeymanmd.org/bio.html.)
A tribute to David Warfield Stires is incomplete without a corresponding acknowledgement and celebration of John’s decades-long visionary leadership in family medicine.
Jay Siwek, MD
Washington, DC
Thank you for your tribute to David Warfield Stires, The Journal of Family Practice’s founding publisher (J Fam Pract. 2017;66:654-655). The real hero of the story, however, is Dr. John Geyman, who had the vision to found a research journal at the birth of our specialty. This was no easy task, as John faced a challenging chicken-and-egg problem: how to establish a research journal when academic family medicine was just getting underway and had no track record of generating a steady stream of quality research. The latter problem was due, in part, to the lack of a research journal devoted to promoting and publishing research in the field.
Yet, John did it, putting family medicine research on the publishing map. His groundbreaking work set the stage for future journals, including the Journal of the American Board of Family Medicine, the American Medical Association’s now-defunct Archives of Family Medicine, and the American Academy of Family Physicians’ Annals of Family Medicine.
As a family medicine resident in the 1970s, I remember coveting JFP so much that I managed to collect every issue from Volume 1, Issue 1, through the turn of the century. And as a young faculty member at Georgetown University Medical Center in the 1980s, I painstakingly created an annotated bibliography of the then-published content of JFP to use for teaching, research, and administration.
When I became editor of American Family Physician in 1988, I made a pilgrimage to the University of Washington School of Medicine in Seattle, where John was chairman of the Department of Family Medicine. I wanted to seek his advice, learn from his vast experience, and pay tribute to all that he’d done for our specialty. Over the past 30 years, John has continued to leave his mark. (See http://www.johngeymanmd.org/bio.html.)
A tribute to David Warfield Stires is incomplete without a corresponding acknowledgement and celebration of John’s decades-long visionary leadership in family medicine.
Jay Siwek, MD
Washington, DC
Utilize guidelines, but customize BP treatment in older patients
In the article, “Hypertension treatment strategies for older adults” (J Fam Pract. 2017;66:546-554), Hansell et al recommend a systolic blood pressure (SBP) treatment target of <120 mm Hg for community-dwelling, nondiabetic patients ≥75 years of age. This recommendation is not supported by the authors’ cited guidelines, and we have serious concerns about the risk of harm from such overly stringent BP control in this population.
While Hansell et al acknowledge that no consensus exists regarding an optimal BP target for older patients, the authors cite the Eighth Joint National Committee (JNC 8), the American College of Physicians (ACP), the Systolic Blood Pressure Intervention Trial (SPRINT) subgroup analysis, and the BP arm of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial to justify their recommendation. But as the authors mention, JNC 8 conducted a comprehensive review of the available evidence and determined that a BP target of <150/90 mm Hg for hypertensive patients ≥60 years of age is appropriate.1
The authors also state that ACP recommends an SBP target of <140 mm Hg, while, in fact, the recommendations from ACP (which are joint guidelines published with the American Academy of Family Physicians) say that high-quality evidence strongly supports an SBP target of <150 mm Hg to reduce the risk for mortality, stroke, and cardiac events in adults ≥60 years of age.2
SPRINT does support Hansell et al’s recommended SBP target of <120 mm Hg, but this trial provided only composite data of adults ≥75 years of age and did not differentiate between the outcomes in otherwise healthy adults ≥75 years of age vs those with cardiovascular conditions.3 As Hansell et al point out, the SPRINT trial was halted prematurely, which compromises the validity of their findings.
Lastly, the ACCORD trial did not find benefit to treating SBP <120 mm Hg compared with <140 mm Hg in adults with diabetes, but it did find substantial harms in the <120 mm Hg group, including an increased risk of renal impairment and hypokalemia.4
Hansel et al’s overreliance on the SPRINT subgroup analysis represents a significant flaw in the assertion that an SBP target <120 mm Hg is reasonable for all community-dwelling, non-diabetic adults ≥75 years of age. While the authors made the allowance that a higher target (<140 mm Hg) is acceptable if a target of <120 mm Hg places undue burden on the patient, the guidelines they cited, when considered together, suggest that starting at a higher target is not only sufficient to prevent complications, but also reduces overtreatment.
Adults ≥75 years of age are a diverse group regarding disease conditions, life expectancy, and personal priorities. While it is tempting to make generalizations about BP treatment targets, we owe it to our patients to understand the nuances of applicable guidelines so that we can tailor BP treatment targets to each patient’s unique clinical situation and personal priorities. Applying a blanket recommendation to this heterogeneous population may result in significant harms from overtreatment.
Jennifer L. Middleton, MD, MPH, FAAFP; Miriam Chan, PharmD, CDE
Columbus, Ohio
1. James PA, Oparil S, Carter BL, et al. 2014 Evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507-520.
2. Qaseem A, Wilt TJ, Rich R, et al. Pharmacological treatment of hypertension in adults aged 60 years or older to higher versus lower blood pressure targets: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2017;166:430-437.
3. Williamson JD, Suplano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
4. ACCORD Study Group, Cushman WC, Evans GW, Byington RP, et al. Effects of intensive blood-pressure control in type 2 diabetes mellitus. N Engl J Med. 2010;362:1575-1585.
Authors’ response:
We agree with the title of this letter, “Utilize guidelines, but customize BP treatment in older patients.” Our recommendations are not limited to targeting a systolic BP <120 mm Hg for community-dwelling, nondiabetic adults ≥75 years of age, but include consideration for “undue burden.” Our third practice recommendation, which recommends that one consider cognitive function, polypharmacy, multimorbidity, and frailty, is an equally—if not more—important recommendation.
With regard to the specific concerns about the current guidelines:
- The American College of Physicians and American Academy of Family Physicians’ “Recommendation 1” advocates a systolic BP goal <150 mm Hg for adults ≥60 years of age. However, “Recommendation 3” endorses intensifying treatment in adults ≥60 years of age at high cardiovascular (CV) risk. Based on Framingham criteria, all adults ≥75 years of age are considered at high risk for CV disease, as stated in our article. Therefore, “Recommendation 3” for a target of <140 mm Hg is applicable for the population addressed in our article.1
- The Eighth Joint National Committee (JNC 8) does recommend a BP target <150 mm Hg for adults ≥60 years of age, but does not take into account recent data, which is why we wanted to highlight that data for physicians.2
- Since submission of our article, The American College of Cardiology/American Heart Association (ACC/AHA) has published its first set of guidelines since 2003, which lowered BP target to <130 mm Hg in patients with high CV risk. Those guidelines outline the validity of SPRINT and the consistency of the existing evidence, including the linear relationship of BP and mortality.3
- SPRINT was halted early specifically because of the mortality benefit in the intensive treatment group, which is ethically appropriate.4 It is unclear to us how this compromises the validity of the trial. There is often concern for bias from early cessation in small trials, but this was a large, well-powered trial.
- The ACC/AHA guidelines also address some of the nuances of ACCORD, which is specific to patients with diabetes (whom we excluded from our first Practice Recommendation). Although no overall mortality benefit was found, there was stroke reduction in this group and additional benefit in the standard glycemia group.3,5 A meta-analysis of SPRINT and ACCORD showed CV disease reduction with a BP target <120 mm Hg.6
Although we do believe that SPRINT is a landmark trial contributing a great deal to our recommendations, we strongly emphasized that comorbidities, frailty, and dementia greatly impact treatment decisions. We stressed that prescribers use caution and slow titration because of adverse effects. Geriatric medicine is a complex art, and one of the goals of our article was to highlight this complexity and emphasize the importance of considering goals of care, comorbidity, frailty, and cognitive function when choosing optimal BP targets.
Maggie W. Hansell, MD; Emily M. Mann, MD; Julienne K. Kirk, PharmD
Winston-Salem, NC
1. Qaseem A, Wilt TJ, Rich R, et al. Pharmacological treatment of hypertension in adults aged 60 years or older to higher versus lower blood pressure targets: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2017;166:430-437.
2. James PA, Oparil S, Carter BL, et al. 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507-520.
3. Whelton PK, Carey RM, Aronow WS, et al. ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Available at: http://hyper.ahajournals.org/content/hypertensionaha/early/2017/11/10/HYP.0000000000000066.full.pdf. Accessed December 12, 2017.
4. SPRINT Research Group, Wright JT Jr, Williamson JD, et al. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2015;373:2103-2106.
5. The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive blood-pressure control in type 2 diabetes. N Engl J Med. 2010;362:1575-1585.
6. Perkovic V, Rodgers A. Redefining blood-pressure targets—SPRINT starts the marathon. N Engl Med. 2015;373:2175-2178.
In the article, “Hypertension treatment strategies for older adults” (J Fam Pract. 2017;66:546-554), Hansell et al recommend a systolic blood pressure (SBP) treatment target of <120 mm Hg for community-dwelling, nondiabetic patients ≥75 years of age. This recommendation is not supported by the authors’ cited guidelines, and we have serious concerns about the risk of harm from such overly stringent BP control in this population.
While Hansell et al acknowledge that no consensus exists regarding an optimal BP target for older patients, the authors cite the Eighth Joint National Committee (JNC 8), the American College of Physicians (ACP), the Systolic Blood Pressure Intervention Trial (SPRINT) subgroup analysis, and the BP arm of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial to justify their recommendation. But as the authors mention, JNC 8 conducted a comprehensive review of the available evidence and determined that a BP target of <150/90 mm Hg for hypertensive patients ≥60 years of age is appropriate.1
The authors also state that ACP recommends an SBP target of <140 mm Hg, while, in fact, the recommendations from ACP (which are joint guidelines published with the American Academy of Family Physicians) say that high-quality evidence strongly supports an SBP target of <150 mm Hg to reduce the risk for mortality, stroke, and cardiac events in adults ≥60 years of age.2
SPRINT does support Hansell et al’s recommended SBP target of <120 mm Hg, but this trial provided only composite data of adults ≥75 years of age and did not differentiate between the outcomes in otherwise healthy adults ≥75 years of age vs those with cardiovascular conditions.3 As Hansell et al point out, the SPRINT trial was halted prematurely, which compromises the validity of their findings.
Lastly, the ACCORD trial did not find benefit to treating SBP <120 mm Hg compared with <140 mm Hg in adults with diabetes, but it did find substantial harms in the <120 mm Hg group, including an increased risk of renal impairment and hypokalemia.4
Hansel et al’s overreliance on the SPRINT subgroup analysis represents a significant flaw in the assertion that an SBP target <120 mm Hg is reasonable for all community-dwelling, non-diabetic adults ≥75 years of age. While the authors made the allowance that a higher target (<140 mm Hg) is acceptable if a target of <120 mm Hg places undue burden on the patient, the guidelines they cited, when considered together, suggest that starting at a higher target is not only sufficient to prevent complications, but also reduces overtreatment.
Adults ≥75 years of age are a diverse group regarding disease conditions, life expectancy, and personal priorities. While it is tempting to make generalizations about BP treatment targets, we owe it to our patients to understand the nuances of applicable guidelines so that we can tailor BP treatment targets to each patient’s unique clinical situation and personal priorities. Applying a blanket recommendation to this heterogeneous population may result in significant harms from overtreatment.
Jennifer L. Middleton, MD, MPH, FAAFP; Miriam Chan, PharmD, CDE
Columbus, Ohio
1. James PA, Oparil S, Carter BL, et al. 2014 Evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507-520.
2. Qaseem A, Wilt TJ, Rich R, et al. Pharmacological treatment of hypertension in adults aged 60 years or older to higher versus lower blood pressure targets: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2017;166:430-437.
3. Williamson JD, Suplano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
4. ACCORD Study Group, Cushman WC, Evans GW, Byington RP, et al. Effects of intensive blood-pressure control in type 2 diabetes mellitus. N Engl J Med. 2010;362:1575-1585.
Authors’ response:
We agree with the title of this letter, “Utilize guidelines, but customize BP treatment in older patients.” Our recommendations are not limited to targeting a systolic BP <120 mm Hg for community-dwelling, nondiabetic adults ≥75 years of age, but include consideration for “undue burden.” Our third practice recommendation, which recommends that one consider cognitive function, polypharmacy, multimorbidity, and frailty, is an equally—if not more—important recommendation.
With regard to the specific concerns about the current guidelines:
- The American College of Physicians and American Academy of Family Physicians’ “Recommendation 1” advocates a systolic BP goal <150 mm Hg for adults ≥60 years of age. However, “Recommendation 3” endorses intensifying treatment in adults ≥60 years of age at high cardiovascular (CV) risk. Based on Framingham criteria, all adults ≥75 years of age are considered at high risk for CV disease, as stated in our article. Therefore, “Recommendation 3” for a target of <140 mm Hg is applicable for the population addressed in our article.1
- The Eighth Joint National Committee (JNC 8) does recommend a BP target <150 mm Hg for adults ≥60 years of age, but does not take into account recent data, which is why we wanted to highlight that data for physicians.2
- Since submission of our article, The American College of Cardiology/American Heart Association (ACC/AHA) has published its first set of guidelines since 2003, which lowered BP target to <130 mm Hg in patients with high CV risk. Those guidelines outline the validity of SPRINT and the consistency of the existing evidence, including the linear relationship of BP and mortality.3
- SPRINT was halted early specifically because of the mortality benefit in the intensive treatment group, which is ethically appropriate.4 It is unclear to us how this compromises the validity of the trial. There is often concern for bias from early cessation in small trials, but this was a large, well-powered trial.
- The ACC/AHA guidelines also address some of the nuances of ACCORD, which is specific to patients with diabetes (whom we excluded from our first Practice Recommendation). Although no overall mortality benefit was found, there was stroke reduction in this group and additional benefit in the standard glycemia group.3,5 A meta-analysis of SPRINT and ACCORD showed CV disease reduction with a BP target <120 mm Hg.6
Although we do believe that SPRINT is a landmark trial contributing a great deal to our recommendations, we strongly emphasized that comorbidities, frailty, and dementia greatly impact treatment decisions. We stressed that prescribers use caution and slow titration because of adverse effects. Geriatric medicine is a complex art, and one of the goals of our article was to highlight this complexity and emphasize the importance of considering goals of care, comorbidity, frailty, and cognitive function when choosing optimal BP targets.
Maggie W. Hansell, MD; Emily M. Mann, MD; Julienne K. Kirk, PharmD
Winston-Salem, NC
1. Qaseem A, Wilt TJ, Rich R, et al. Pharmacological treatment of hypertension in adults aged 60 years or older to higher versus lower blood pressure targets: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2017;166:430-437.
2. James PA, Oparil S, Carter BL, et al. 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507-520.
3. Whelton PK, Carey RM, Aronow WS, et al. ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Available at: http://hyper.ahajournals.org/content/hypertensionaha/early/2017/11/10/HYP.0000000000000066.full.pdf. Accessed December 12, 2017.
4. SPRINT Research Group, Wright JT Jr, Williamson JD, et al. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2015;373:2103-2106.
5. The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive blood-pressure control in type 2 diabetes. N Engl J Med. 2010;362:1575-1585.
6. Perkovic V, Rodgers A. Redefining blood-pressure targets—SPRINT starts the marathon. N Engl Med. 2015;373:2175-2178.
In the article, “Hypertension treatment strategies for older adults” (J Fam Pract. 2017;66:546-554), Hansell et al recommend a systolic blood pressure (SBP) treatment target of <120 mm Hg for community-dwelling, nondiabetic patients ≥75 years of age. This recommendation is not supported by the authors’ cited guidelines, and we have serious concerns about the risk of harm from such overly stringent BP control in this population.
While Hansell et al acknowledge that no consensus exists regarding an optimal BP target for older patients, the authors cite the Eighth Joint National Committee (JNC 8), the American College of Physicians (ACP), the Systolic Blood Pressure Intervention Trial (SPRINT) subgroup analysis, and the BP arm of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial to justify their recommendation. But as the authors mention, JNC 8 conducted a comprehensive review of the available evidence and determined that a BP target of <150/90 mm Hg for hypertensive patients ≥60 years of age is appropriate.1
The authors also state that ACP recommends an SBP target of <140 mm Hg, while, in fact, the recommendations from ACP (which are joint guidelines published with the American Academy of Family Physicians) say that high-quality evidence strongly supports an SBP target of <150 mm Hg to reduce the risk for mortality, stroke, and cardiac events in adults ≥60 years of age.2
SPRINT does support Hansell et al’s recommended SBP target of <120 mm Hg, but this trial provided only composite data of adults ≥75 years of age and did not differentiate between the outcomes in otherwise healthy adults ≥75 years of age vs those with cardiovascular conditions.3 As Hansell et al point out, the SPRINT trial was halted prematurely, which compromises the validity of their findings.
Lastly, the ACCORD trial did not find benefit to treating SBP <120 mm Hg compared with <140 mm Hg in adults with diabetes, but it did find substantial harms in the <120 mm Hg group, including an increased risk of renal impairment and hypokalemia.4
Hansel et al’s overreliance on the SPRINT subgroup analysis represents a significant flaw in the assertion that an SBP target <120 mm Hg is reasonable for all community-dwelling, non-diabetic adults ≥75 years of age. While the authors made the allowance that a higher target (<140 mm Hg) is acceptable if a target of <120 mm Hg places undue burden on the patient, the guidelines they cited, when considered together, suggest that starting at a higher target is not only sufficient to prevent complications, but also reduces overtreatment.
Adults ≥75 years of age are a diverse group regarding disease conditions, life expectancy, and personal priorities. While it is tempting to make generalizations about BP treatment targets, we owe it to our patients to understand the nuances of applicable guidelines so that we can tailor BP treatment targets to each patient’s unique clinical situation and personal priorities. Applying a blanket recommendation to this heterogeneous population may result in significant harms from overtreatment.
Jennifer L. Middleton, MD, MPH, FAAFP; Miriam Chan, PharmD, CDE
Columbus, Ohio
1. James PA, Oparil S, Carter BL, et al. 2014 Evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507-520.
2. Qaseem A, Wilt TJ, Rich R, et al. Pharmacological treatment of hypertension in adults aged 60 years or older to higher versus lower blood pressure targets: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2017;166:430-437.
3. Williamson JD, Suplano MA, Applegate WB, et al. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial. JAMA. 2016;315:2673-2682.
4. ACCORD Study Group, Cushman WC, Evans GW, Byington RP, et al. Effects of intensive blood-pressure control in type 2 diabetes mellitus. N Engl J Med. 2010;362:1575-1585.
Authors’ response:
We agree with the title of this letter, “Utilize guidelines, but customize BP treatment in older patients.” Our recommendations are not limited to targeting a systolic BP <120 mm Hg for community-dwelling, nondiabetic adults ≥75 years of age, but include consideration for “undue burden.” Our third practice recommendation, which recommends that one consider cognitive function, polypharmacy, multimorbidity, and frailty, is an equally—if not more—important recommendation.
With regard to the specific concerns about the current guidelines:
- The American College of Physicians and American Academy of Family Physicians’ “Recommendation 1” advocates a systolic BP goal <150 mm Hg for adults ≥60 years of age. However, “Recommendation 3” endorses intensifying treatment in adults ≥60 years of age at high cardiovascular (CV) risk. Based on Framingham criteria, all adults ≥75 years of age are considered at high risk for CV disease, as stated in our article. Therefore, “Recommendation 3” for a target of <140 mm Hg is applicable for the population addressed in our article.1
- The Eighth Joint National Committee (JNC 8) does recommend a BP target <150 mm Hg for adults ≥60 years of age, but does not take into account recent data, which is why we wanted to highlight that data for physicians.2
- Since submission of our article, The American College of Cardiology/American Heart Association (ACC/AHA) has published its first set of guidelines since 2003, which lowered BP target to <130 mm Hg in patients with high CV risk. Those guidelines outline the validity of SPRINT and the consistency of the existing evidence, including the linear relationship of BP and mortality.3
- SPRINT was halted early specifically because of the mortality benefit in the intensive treatment group, which is ethically appropriate.4 It is unclear to us how this compromises the validity of the trial. There is often concern for bias from early cessation in small trials, but this was a large, well-powered trial.
- The ACC/AHA guidelines also address some of the nuances of ACCORD, which is specific to patients with diabetes (whom we excluded from our first Practice Recommendation). Although no overall mortality benefit was found, there was stroke reduction in this group and additional benefit in the standard glycemia group.3,5 A meta-analysis of SPRINT and ACCORD showed CV disease reduction with a BP target <120 mm Hg.6
Although we do believe that SPRINT is a landmark trial contributing a great deal to our recommendations, we strongly emphasized that comorbidities, frailty, and dementia greatly impact treatment decisions. We stressed that prescribers use caution and slow titration because of adverse effects. Geriatric medicine is a complex art, and one of the goals of our article was to highlight this complexity and emphasize the importance of considering goals of care, comorbidity, frailty, and cognitive function when choosing optimal BP targets.
Maggie W. Hansell, MD; Emily M. Mann, MD; Julienne K. Kirk, PharmD
Winston-Salem, NC
1. Qaseem A, Wilt TJ, Rich R, et al. Pharmacological treatment of hypertension in adults aged 60 years or older to higher versus lower blood pressure targets: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2017;166:430-437.
2. James PA, Oparil S, Carter BL, et al. 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507-520.
3. Whelton PK, Carey RM, Aronow WS, et al. ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Available at: http://hyper.ahajournals.org/content/hypertensionaha/early/2017/11/10/HYP.0000000000000066.full.pdf. Accessed December 12, 2017.
4. SPRINT Research Group, Wright JT Jr, Williamson JD, et al. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2015;373:2103-2106.
5. The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive blood-pressure control in type 2 diabetes. N Engl J Med. 2010;362:1575-1585.
6. Perkovic V, Rodgers A. Redefining blood-pressure targets—SPRINT starts the marathon. N Engl Med. 2015;373:2175-2178.
Test storyHeadline – 7 words, sentence case, active verb, include best SEO terms
Body text starts here….
Second and subsequent grafs
Doctor’s Name and Bio
Body text starts here….
Second and subsequent grafs
Doctor’s Name and Bio
Body text starts here….
Second and subsequent grafs
Doctor’s Name and Bio
Key clinical point:
Major finding: Key numerical finding (e.g., number needed to treat to prevent one death/event; number lived or died as result of intervention). Maximum 10 words/1 sentence.
Data source: Include type of study (e.g., randomized, placebo controlled trial; retrospective case-control study). Include number in the study.
Disclosures: Sponsor of study, funding source, relevant disclosures. If author has no relevant disclosures, “Dr. X reported having no financial disclosures.” If necessary, “Meeting Y did not require reports of financial disclosures.” Check meeting website because many list disclosures. Written in sentence form.
Fetal alcohol spectrum disorders and suicidality
As psychiatrists, we understand that behavior is complex and determined by multiple factors. However, despite our understanding that behavior is cultural, sociological, psychological, and biological, we often lose sight of the biological perspective because the brain is such a complex organ and because we are inundated with psychological theories of behavior. As I have said before, we cannot abdicate our role of being biologists in the reflection of mental health and wellness.
Accordingly, I feel it is my duty to bring our attention to a biologic etiology of suicidal behavior. I came across an article on the life expectancy of individuals afflicted with fetal alcohol syndrome in the Journal of Population Therapeutics and Clinical Pharmacology (2016;23[1]:e53-9). The findings were astonishing. As it turns out, the life expectancy of people with fetal alcohol syndrome is 34 years of age on average, and the leading causes of death were “external causes,” which accounted for 44% of the deaths. Suicide was responsible for 15% of those deaths, accidents for 14%, poisoning by illegal drugs or alcohol for 7%, and other external causes for another 7%, according to the article.
While working in a general hospital in a low-income African American environment where there are high rates of fetal alcohol exposure, I see at least 3-4 suicide attempts a week on the medical-surgical/psychiatric inpatient units where I serve. I am always looking for patients who have ND-PAE because determining such a diagnosis is critical to those patients’ medical-surgical care. For example, there was one woman with ND-PAE who had operable breast carcinoma but did not come in for a return visit until after her carcinoma had become inoperable (she forgot how important it was to get timely treatment). There was a patient who always had out-of-control diabetes because he did not know how to use his glucometer. There was a patient who was taking his antipsychotic medication during the day instead of as prescribed – at bedtime – because he could not read the instructions on his medication bottle. (I have altered several key aspects of my patients’ stories to protect their confidentiality.)
However, until I read that suicide was responsible for 15% of deaths with external causes among patients with fetal alcohol syndrome – patients whose life expectancy averages only 34 years – it did not occur to me that affect dysregulation also was likely to lead to suicide attempts among patients with ND-PAE.
When several of us who were working on the issue of suicide prevention while part of the Committee on Psychopathology & Prevention of Adolescent & Adult Suicide produced our report called “Reducing Suicide: A National Perspective” in 2002, the idea that paying attention to fetal environments and birth outcomes could inform the area of suicide prevention was an alien one. Now, it is a serious consideration because this dynamic just might explain part of the complex phenomena of some suicidal behaviors.
Dr. Bell is a staff psychiatrist at Jackson Park Hospital Family Medicine Clinic in Chicago; clinical psychiatrist emeritus in the department of psychiatry at the University of Illinois at Chicago; former president/CEO of Community Mental Health Council; and former director of the Institute for Juvenile Research (birthplace of child psychiatry), also in Chicago.
As psychiatrists, we understand that behavior is complex and determined by multiple factors. However, despite our understanding that behavior is cultural, sociological, psychological, and biological, we often lose sight of the biological perspective because the brain is such a complex organ and because we are inundated with psychological theories of behavior. As I have said before, we cannot abdicate our role of being biologists in the reflection of mental health and wellness.
Accordingly, I feel it is my duty to bring our attention to a biologic etiology of suicidal behavior. I came across an article on the life expectancy of individuals afflicted with fetal alcohol syndrome in the Journal of Population Therapeutics and Clinical Pharmacology (2016;23[1]:e53-9). The findings were astonishing. As it turns out, the life expectancy of people with fetal alcohol syndrome is 34 years of age on average, and the leading causes of death were “external causes,” which accounted for 44% of the deaths. Suicide was responsible for 15% of those deaths, accidents for 14%, poisoning by illegal drugs or alcohol for 7%, and other external causes for another 7%, according to the article.
While working in a general hospital in a low-income African American environment where there are high rates of fetal alcohol exposure, I see at least 3-4 suicide attempts a week on the medical-surgical/psychiatric inpatient units where I serve. I am always looking for patients who have ND-PAE because determining such a diagnosis is critical to those patients’ medical-surgical care. For example, there was one woman with ND-PAE who had operable breast carcinoma but did not come in for a return visit until after her carcinoma had become inoperable (she forgot how important it was to get timely treatment). There was a patient who always had out-of-control diabetes because he did not know how to use his glucometer. There was a patient who was taking his antipsychotic medication during the day instead of as prescribed – at bedtime – because he could not read the instructions on his medication bottle. (I have altered several key aspects of my patients’ stories to protect their confidentiality.)
However, until I read that suicide was responsible for 15% of deaths with external causes among patients with fetal alcohol syndrome – patients whose life expectancy averages only 34 years – it did not occur to me that affect dysregulation also was likely to lead to suicide attempts among patients with ND-PAE.
When several of us who were working on the issue of suicide prevention while part of the Committee on Psychopathology & Prevention of Adolescent & Adult Suicide produced our report called “Reducing Suicide: A National Perspective” in 2002, the idea that paying attention to fetal environments and birth outcomes could inform the area of suicide prevention was an alien one. Now, it is a serious consideration because this dynamic just might explain part of the complex phenomena of some suicidal behaviors.
Dr. Bell is a staff psychiatrist at Jackson Park Hospital Family Medicine Clinic in Chicago; clinical psychiatrist emeritus in the department of psychiatry at the University of Illinois at Chicago; former president/CEO of Community Mental Health Council; and former director of the Institute for Juvenile Research (birthplace of child psychiatry), also in Chicago.
As psychiatrists, we understand that behavior is complex and determined by multiple factors. However, despite our understanding that behavior is cultural, sociological, psychological, and biological, we often lose sight of the biological perspective because the brain is such a complex organ and because we are inundated with psychological theories of behavior. As I have said before, we cannot abdicate our role of being biologists in the reflection of mental health and wellness.
Accordingly, I feel it is my duty to bring our attention to a biologic etiology of suicidal behavior. I came across an article on the life expectancy of individuals afflicted with fetal alcohol syndrome in the Journal of Population Therapeutics and Clinical Pharmacology (2016;23[1]:e53-9). The findings were astonishing. As it turns out, the life expectancy of people with fetal alcohol syndrome is 34 years of age on average, and the leading causes of death were “external causes,” which accounted for 44% of the deaths. Suicide was responsible for 15% of those deaths, accidents for 14%, poisoning by illegal drugs or alcohol for 7%, and other external causes for another 7%, according to the article.
While working in a general hospital in a low-income African American environment where there are high rates of fetal alcohol exposure, I see at least 3-4 suicide attempts a week on the medical-surgical/psychiatric inpatient units where I serve. I am always looking for patients who have ND-PAE because determining such a diagnosis is critical to those patients’ medical-surgical care. For example, there was one woman with ND-PAE who had operable breast carcinoma but did not come in for a return visit until after her carcinoma had become inoperable (she forgot how important it was to get timely treatment). There was a patient who always had out-of-control diabetes because he did not know how to use his glucometer. There was a patient who was taking his antipsychotic medication during the day instead of as prescribed – at bedtime – because he could not read the instructions on his medication bottle. (I have altered several key aspects of my patients’ stories to protect their confidentiality.)
However, until I read that suicide was responsible for 15% of deaths with external causes among patients with fetal alcohol syndrome – patients whose life expectancy averages only 34 years – it did not occur to me that affect dysregulation also was likely to lead to suicide attempts among patients with ND-PAE.
When several of us who were working on the issue of suicide prevention while part of the Committee on Psychopathology & Prevention of Adolescent & Adult Suicide produced our report called “Reducing Suicide: A National Perspective” in 2002, the idea that paying attention to fetal environments and birth outcomes could inform the area of suicide prevention was an alien one. Now, it is a serious consideration because this dynamic just might explain part of the complex phenomena of some suicidal behaviors.
Dr. Bell is a staff psychiatrist at Jackson Park Hospital Family Medicine Clinic in Chicago; clinical psychiatrist emeritus in the department of psychiatry at the University of Illinois at Chicago; former president/CEO of Community Mental Health Council; and former director of the Institute for Juvenile Research (birthplace of child psychiatry), also in Chicago.
Identity crisis
The provider has received “advanced-level education in pharmacology, pathophysiology, and physical assessment, diagnosis, and management” and provides patient care in a medical home “in a holistic fashion including physical care, therapeutic treatments, education, and coordination of services.”
This quote comes from a recent story in Pediatric News about collaborative practice. Was the author offering a job description of a) a chiropractor, b) a nurse practitioner, c) a pediatric oncologist, or d) a primary care physician?
Based on my personal experience working with nurse practitioners, both in hospital and office settings, I wholeheartedly concur with Dr. Haut’s list of their qualifications and capabilities. My problem is that she doesn’t list, nor can I comfortably imagine, the additional skills that a physician should have in his or her toolbox to complete the complementary relationships in a primary care practice that Dr. Haut envisions.
From my perspective, nurse practitioners and primary care physicians share the same job description, the one I listed in the first paragraph of this column. They both provide face-to-face, usually hands-on, medical care. At that critical interface between patient and provider, how do their roles differ? What other skills does a physician need to complement those of a competent and already experienced nurse practitioner?
Does being a physician guarantee that he or she has more experience than a nurse practitioner? You know as well as I do that you finished your training pretty wet behind the ears, and the first 5 years or more of your practice career were when you really began to feel like a competent provider. If my child has an earache, I would probably be more comfortable, or at least as comfortable, with her seeing a nurse practitioner with 5 years of experience in a busy practice than a newly minted, board-eligible pediatrician.
Is the breadth of a physician’s training in medical school an asset? Does the 2-month rotation he or she did on the adult neurology service taking care of stroke victims give the physician an advantage when it comes to taking care of pediatric patients with asthma?
Actually, I can imagine a suite of skills that a physician might bring to a collaborative practice that a nurse practitioner may not have, or more likely may have chosen not to pursue. Those skills have little to do with direct patient care, but can be critical for survival in today’s medical care environment. Here I am thinking of things such as negotiating with third-party payers, and leading and/or administering the complexities of a medium-sized or larger medical group. Does having a degree from a medical school automatically mean that the graduate is a skilled leader or administrator?
I can envision that over time a physician and a nurse practitioner might create an arrangement in which one of them focuses on the patients with asthma and attention-deficit/hyperactivity disorder, and the other develops an expertise in breastfeeding management and picky eating. That kind of relationship fits my definition of complementary. However, a relationship in which the doctor is the boss and the nurse practitioner is not doesn’t feel complementary or collaborative to me.
Dr. Wilkoff practiced primary care pediatrics in Brunswick, Maine, for nearly 40 years. He has authored several books on behavioral pediatrics, including “How to Say No to Your Toddler.”
The provider has received “advanced-level education in pharmacology, pathophysiology, and physical assessment, diagnosis, and management” and provides patient care in a medical home “in a holistic fashion including physical care, therapeutic treatments, education, and coordination of services.”
This quote comes from a recent story in Pediatric News about collaborative practice. Was the author offering a job description of a) a chiropractor, b) a nurse practitioner, c) a pediatric oncologist, or d) a primary care physician?
Based on my personal experience working with nurse practitioners, both in hospital and office settings, I wholeheartedly concur with Dr. Haut’s list of their qualifications and capabilities. My problem is that she doesn’t list, nor can I comfortably imagine, the additional skills that a physician should have in his or her toolbox to complete the complementary relationships in a primary care practice that Dr. Haut envisions.
From my perspective, nurse practitioners and primary care physicians share the same job description, the one I listed in the first paragraph of this column. They both provide face-to-face, usually hands-on, medical care. At that critical interface between patient and provider, how do their roles differ? What other skills does a physician need to complement those of a competent and already experienced nurse practitioner?
Does being a physician guarantee that he or she has more experience than a nurse practitioner? You know as well as I do that you finished your training pretty wet behind the ears, and the first 5 years or more of your practice career were when you really began to feel like a competent provider. If my child has an earache, I would probably be more comfortable, or at least as comfortable, with her seeing a nurse practitioner with 5 years of experience in a busy practice than a newly minted, board-eligible pediatrician.
Is the breadth of a physician’s training in medical school an asset? Does the 2-month rotation he or she did on the adult neurology service taking care of stroke victims give the physician an advantage when it comes to taking care of pediatric patients with asthma?
Actually, I can imagine a suite of skills that a physician might bring to a collaborative practice that a nurse practitioner may not have, or more likely may have chosen not to pursue. Those skills have little to do with direct patient care, but can be critical for survival in today’s medical care environment. Here I am thinking of things such as negotiating with third-party payers, and leading and/or administering the complexities of a medium-sized or larger medical group. Does having a degree from a medical school automatically mean that the graduate is a skilled leader or administrator?
I can envision that over time a physician and a nurse practitioner might create an arrangement in which one of them focuses on the patients with asthma and attention-deficit/hyperactivity disorder, and the other develops an expertise in breastfeeding management and picky eating. That kind of relationship fits my definition of complementary. However, a relationship in which the doctor is the boss and the nurse practitioner is not doesn’t feel complementary or collaborative to me.
Dr. Wilkoff practiced primary care pediatrics in Brunswick, Maine, for nearly 40 years. He has authored several books on behavioral pediatrics, including “How to Say No to Your Toddler.”
The provider has received “advanced-level education in pharmacology, pathophysiology, and physical assessment, diagnosis, and management” and provides patient care in a medical home “in a holistic fashion including physical care, therapeutic treatments, education, and coordination of services.”
This quote comes from a recent story in Pediatric News about collaborative practice. Was the author offering a job description of a) a chiropractor, b) a nurse practitioner, c) a pediatric oncologist, or d) a primary care physician?
Based on my personal experience working with nurse practitioners, both in hospital and office settings, I wholeheartedly concur with Dr. Haut’s list of their qualifications and capabilities. My problem is that she doesn’t list, nor can I comfortably imagine, the additional skills that a physician should have in his or her toolbox to complete the complementary relationships in a primary care practice that Dr. Haut envisions.
From my perspective, nurse practitioners and primary care physicians share the same job description, the one I listed in the first paragraph of this column. They both provide face-to-face, usually hands-on, medical care. At that critical interface between patient and provider, how do their roles differ? What other skills does a physician need to complement those of a competent and already experienced nurse practitioner?
Does being a physician guarantee that he or she has more experience than a nurse practitioner? You know as well as I do that you finished your training pretty wet behind the ears, and the first 5 years or more of your practice career were when you really began to feel like a competent provider. If my child has an earache, I would probably be more comfortable, or at least as comfortable, with her seeing a nurse practitioner with 5 years of experience in a busy practice than a newly minted, board-eligible pediatrician.
Is the breadth of a physician’s training in medical school an asset? Does the 2-month rotation he or she did on the adult neurology service taking care of stroke victims give the physician an advantage when it comes to taking care of pediatric patients with asthma?
Actually, I can imagine a suite of skills that a physician might bring to a collaborative practice that a nurse practitioner may not have, or more likely may have chosen not to pursue. Those skills have little to do with direct patient care, but can be critical for survival in today’s medical care environment. Here I am thinking of things such as negotiating with third-party payers, and leading and/or administering the complexities of a medium-sized or larger medical group. Does having a degree from a medical school automatically mean that the graduate is a skilled leader or administrator?
I can envision that over time a physician and a nurse practitioner might create an arrangement in which one of them focuses on the patients with asthma and attention-deficit/hyperactivity disorder, and the other develops an expertise in breastfeeding management and picky eating. That kind of relationship fits my definition of complementary. However, a relationship in which the doctor is the boss and the nurse practitioner is not doesn’t feel complementary or collaborative to me.
Dr. Wilkoff practiced primary care pediatrics in Brunswick, Maine, for nearly 40 years. He has authored several books on behavioral pediatrics, including “How to Say No to Your Toddler.”