Letters
tomatic patients may help guide clinical decision making and optimal clinical practice. Kosmas I. Paraskevas, MD Frank J. Veith, MD Author Affiliations: Department of Vascular Surgery, Larissa University Hospital, Larissa, Greece (Paraskevas); Division of Vascular Surgery, New York University Medical Center, New York (Veith). Corresponding Author: Kosmas I. Paraskevas, MD, Department of Vascular Surgery, Larissa University Hospital, Larissa 41100, Greece (paraskevask @hotmail.com). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Beckman JA. Management of asymptomatic internal carotid artery stenosis. JAMA. 2013;310(15):1612-1618. 2. Brott TG, Hobson RW II, Howard G, et al; CREST Investigators. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med. 2010;363(1):11-23.
strokes are rare with both types of revascularization, particularly those that cause persistent symptoms. The difference between modalities is driven by minor time-limited events without persistent effect. Third, the meta-analysis cited by Paraskevas and Veith combined clinical trials of both symptomatic and asymptomatic patients. Symptomatic patients have a significantly higher stroke risk and different pathophysiology (heightened local and systemic inflammation and tendency toward thrombosis3-6), rendering conclusions derived from a combination of data from symptomatic and asymptomatic patients inappropriate. As noted in the Grand Rounds article, none of the 4 clinical trials comparing CAS with CEA that included asymptomatic patients demonstrated superiority for either modality in patients with average risk. Joshua A. Beckman, MD, MS
3. Howard VJ, Lutsep HL, Mackey A, et al; CREST Investigators. Influence of sex on outcomes of stenting versus endarterectomy: a subgroup analysis of the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST). Lancet Neurol. 2011;10(6):530-537.
Author Affiliation: Cardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts.
4. Voeks JH, Howard G, Roubin GS, et al; CREST Investigators. Age and outcomes after carotid stenting and endarterectomy: the carotid revascularization endarterectomy versus stenting trial. Stroke. 2011;42(12):3484-3490.
Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and reported being a board member of Vascular Interventional Advances; receiving grant funding from Bristol-Myers Squibb; and being a consultant for Novartis, Ferring, Merck, Boston Scientific, and AstraZeneca.
5. Antoniou GA, Georgiadis GS, Georgakarakos EI, et al. Meta-analysis and meta-regression analysis of outcomes of carotid endarterectomy and stenting in the elderly. JAMA Surg. 2013;148(12):1140-1152.
In Reply Drs Paraskevas and Veith discuss CREST subgroup analyses1 to suggest CEA would have had a lower stroke risk than CAS in the patient presented in the Grand Rounds article. In each study, the subgroup analysis was done using the factor of interest (age or sex), but these treatment-by-factor relationships were not affected by symptomatic status. Their assessment is mitigated by 3 factors: (1) lack of statistical power, (2) the treatment of stroke as an event of singular severity, and (3) the use of data that combines symptomatic and asymptomatic patients. First, Paraskevas and Veith report that symptomatic status does not affect the outcome; however, that would mistake proof for lack of statistical power. As noted in the substudy of CREST by symptomatic status,2 the total number of strokes among the 1181 asymptomatic participants was 23 (15 receiving CAS and 8 receiving CEA). The event rate was low and the difference was not significant, with the HRs ranging from 1.50 to 2.06. Any attempt to analyze additional subgroups, whether by age or sex, would only make the event total smaller, further limiting the statistical power to understand the effect of age or sex in asymptomatic patients. Second, Paraskevas and Veith describe stroke as though it is a singular item. In CREST, both minor and major stroke events are reported.2 Minor strokes, which accounted for 18 of the 23 total stroke events, were defined as those in which patients were asymptomatic by 30 days. Major stroke, which accounted for 5 total events in the 1181 patient cohort, was defined as a stroke with 9 or more points at 3 months on the National Institutes of Health Stroke Scale. CREST shows that
Corresponding Author: Joshua A. Beckman, MD, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115 ([email protected]).
1. Brott TG, Hobson RW II, Howard G, et al; CREST Investigators. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med. 2010;363(1):11-23. 2. Silver FL, Mackey A, Clark WM, et al; CREST Investigators. Safety of stenting and endarterectomy by symptomatic status in the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST). Stroke. 2011;42(3):675-680. 3. Ritter MA, Jurk K, Schriek C, et al. Microembolic signals on transcranial Doppler ultrasound are correlated with platelet activation markers, but not with platelet-leukocyte associates: a study in patients with acute stroke and in patients with asymptomatic carotid stenosis. Neurol Res. 2009;31(1):11-16. 4. Kinsella JA, Tobin WO, Cox D, et al. Prevalence of ex vivo high on-treatment platelet reactivity on antiplatelet therapy after transient ischemic attack or ischemic stroke on the PFA-100(®) and VerifyNow(®). J Stroke Cerebrovasc Dis. 2013;22(7):e84-e92. 5. Abbas A, Aukrust P, Dahl TB, et al. High levels of S100A12 are associated with recent plaque symptomatology in patients with carotid atherosclerosis [published correction appears in Stroke. 2013;44(7):e82]. Stroke. 2012;43(5):1347-1353. 6. Koutouzis M, Rallidis LS, Peros G, et al. Serum interleukin-6 is elevated in symptomatic carotid bifurcation disease. Acta Neurol Scand. 2009;119(2):119-125.
Diagnosis of Neonatal Infection With Herpes Simplex Virus To the Editor In the JAMA Clinical Challenge presenting an infant with neonatal infection with herpes simplex virus (HSV), Dr Ladizinski and colleagues1 recommended performing rapid direct fluorescence antibody (DFA) testing from a vesicle to establish the diagnosis. Ladizinski et al1 stated that DFA should be chosen over alternative diagnostic techniques, including viral cultures and polymerase chain reaction (PCR) assays, because DFA provides results most rapidly. However, recent recommendations from both the Infectious Diseases Society of America and the American Society for Microbiology characterize nucleic acid amplification test-
jama.com
JAMA February 5, 2014 Volume 311, Number 5
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of St. Andrews Library User on 06/07/2015
527
Letters
ing, such as PCR, as the “most sensitive, specific, and rapid test for diagnosis of HSV-associated skin or mucosal lesions.”2 These characteristics of PCR explain why our institution’s laboratory no longer offers in-house DFA or viral culture as techniques to diagnose cutaneous HSV. Obtaining a specimen for PCR is not difficult; one simply scrapes the base of a cutaneous or mucosal lesion using a swab.2 A PCR assay also has the ability to differentiate between HSV types 1 and 2.2 Additionally, the most recent guidelines from the US Centers for Disease Control and Prevention omit DFA entirely from their discussion on the evaluation of possible HSV infection.3 Instead, they recommend PCR or viral culture as the preferred tests, adding that the sensitivity of PCR is superior to that of culture.3 Thus PCR, rather than DFA, is the diagnostic test of choice. The physician’s decision regarding which diagnostic test to order depends, ultimately, on the laboratory capabilities of the institution, specifically which tests are available and how frequently the tests are performed. Few hospital laboratories run HSV diagnostic tests on a per case basis and most generally perform (or send out) the tests infrequently during the work week and not on weekends at all. Under these circumstances, consultation with the clinical laboratory can help determine which test will provide the answer soonest. All else being equal, it seems that PCR should be regarded as the procedure of choice to diagnose neonatal HSV. Antiviral therapy should be initiated while waiting for the test results in a case of suspected neonatal HSV. Nathan W. Rojek, BS Scott A. Norton, MD, MPH Author Affiliations: Georgetown University School of Medicine, Washington, DC (Rojek); Department of Dermatology, Children’s National Medical Center, Washington, DC (Norton). Corresponding Author: Nathan W. Rojek, BS, Georgetown University School of Medicine, 2456 Tunlaw Rd NW, Washington, DC 20007 ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Ladizinski B, Rukhman E, Lee KC. A 4-day-old neonate with a widespread vesicular rash. JAMA. 2013;310(9):971-972. 2. Baron EJ, Miller JM, Weinstein MP, et al. A guide to utilization of the microbiology laboratory for diagnosis of infectious diseases: 2013 recommendations by the Infectious Diseases Society of America (IDSA) and the American Society for Microbiology (ASM)(a). Clin Infect Dis. 2013;57(4):e22e121. 3. Centers for Disease Control and Prevention. Diseases characterized by genital, anal, or perianal ulcers. http://www.cdc.gov/std/treatment/2010 /genital-ulcers.htm#hsv. Accessed September 6, 2013.
In Reply Even though rapid DFA testing is not the reference standard test for diagnosing HSV, it was the best option among those given in the answer choices, and the most rapid diagnostic test available because PCR was not an option at the hospital during the weekend when the infant presented.
528
Given that PCR is the preferred diagnostic method based on current recommendations,1,2 it was ordered during the week. Meanwhile, the rapid DFA test was sent due to the availability of the test. This case highlights the utility of additional testing methods for diagnosis of neonatal HSV. Barry Ladizinski, MD Kachiu C. Lee, MD, MPH Author Affiliations: Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland (Ladizinski); Brown University, Providence, Rhode Island (Lee). Corresponding Author: Barry Ladizinski, MD, Johns Hopkins Bloomberg School of Public Health, 300 N Charles St, Ste 304, Baltimore, MD 21201 ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Centers for Disease Control and Prevention. Diseases characterized by genital, anal, or perianal ulcers. http://www.cdc.gov/std/treatment/2010 /genital-ulcers.htm#hsv. Accessed October 7, 2013. 2. Baron EJ, Miller JM, Weinstein MP, et al. A guide to utilization of the microbiology laboratory for diagnosis of infectious diseases: 2013 recommendations by the Infectious Diseases Society of America (IDSA) and the American Society for Microbiology (ASM)(a). Clin Infect Dis. 2013;57(4):e22-e121.
CORRECTION Patient Data Switched in Table: In the Original Investigation titled “Risk of Major Adverse Cardiac Events Following Noncardiac Surgery in Patients With Coronary Stents,” published in the October 9, 2013, issue of JAMA (2013;310[14]:14621472. doi:10.1001/jama.2013.278787), the numbers of male and female patients were switched in the first column of data in Table 1. This article has been corrected online.
Guidelines for Letters Letters discussing a recent JAMA article should be submitted within 4 weeks of the article's publication in print. Letters received after 4 weeks will rarely be considered. Letters should not exceed 400 words of text and 5 references and may have no more than 3 authors. Letters reporting original research should not exceed 600 words of text and 6 references and may have no more than 5 authors. They may include up to 2 tables or figures but online supplementary material is not allowed. All letters should include a word count. Letters must not duplicate other material published or submitted for publication. Letters not meeting these specifications are generally not considered. Letters being considered for publication ordinarily will be sent to the authors of the JAMA article, who will be given the opportunity to reply. Letters will be published at the discretion of the editors and are subject to abridgement and editing. Further instructions can be found at http://jama.com/public /InstructionsForAuthors.aspx. A signed statement for authorship criteria and responsibility, financial disclosure, copyright transfer, and acknowledgment and the ICMJE Form for Disclosure of Potential Conflicts of Interest are required before publication. Letters should be submitted via the JAMA online submission and review system at http: //manuscripts.jama.com. For technical assistance, please contact [email protected]. Section Editor: Jody W. Zylke, MD, Senior Editor.
JAMA February 5, 2014 Volume 311, Number 5
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of St. Andrews Library User on 06/07/2015
jama.com
tomatic patients may help guide clinical decision making and optimal clinical practice. Kosmas I. Paraskevas, MD Frank J. Veith, MD Author Affiliations: Department of Vascular Surgery, Larissa University Hospital, Larissa, Greece (Paraskevas); Division of Vascular Surgery, New York University Medical Center, New York (Veith). Corresponding Author: Kosmas I. Paraskevas, MD, Department of Vascular Surgery, Larissa University Hospital, Larissa 41100, Greece (paraskevask @hotmail.com). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Beckman JA. Management of asymptomatic internal carotid artery stenosis. JAMA. 2013;310(15):1612-1618. 2. Brott TG, Hobson RW II, Howard G, et al; CREST Investigators. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med. 2010;363(1):11-23.
strokes are rare with both types of revascularization, particularly those that cause persistent symptoms. The difference between modalities is driven by minor time-limited events without persistent effect. Third, the meta-analysis cited by Paraskevas and Veith combined clinical trials of both symptomatic and asymptomatic patients. Symptomatic patients have a significantly higher stroke risk and different pathophysiology (heightened local and systemic inflammation and tendency toward thrombosis3-6), rendering conclusions derived from a combination of data from symptomatic and asymptomatic patients inappropriate. As noted in the Grand Rounds article, none of the 4 clinical trials comparing CAS with CEA that included asymptomatic patients demonstrated superiority for either modality in patients with average risk. Joshua A. Beckman, MD, MS
3. Howard VJ, Lutsep HL, Mackey A, et al; CREST Investigators. Influence of sex on outcomes of stenting versus endarterectomy: a subgroup analysis of the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST). Lancet Neurol. 2011;10(6):530-537.
Author Affiliation: Cardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts.
4. Voeks JH, Howard G, Roubin GS, et al; CREST Investigators. Age and outcomes after carotid stenting and endarterectomy: the carotid revascularization endarterectomy versus stenting trial. Stroke. 2011;42(12):3484-3490.
Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and reported being a board member of Vascular Interventional Advances; receiving grant funding from Bristol-Myers Squibb; and being a consultant for Novartis, Ferring, Merck, Boston Scientific, and AstraZeneca.
5. Antoniou GA, Georgiadis GS, Georgakarakos EI, et al. Meta-analysis and meta-regression analysis of outcomes of carotid endarterectomy and stenting in the elderly. JAMA Surg. 2013;148(12):1140-1152.
In Reply Drs Paraskevas and Veith discuss CREST subgroup analyses1 to suggest CEA would have had a lower stroke risk than CAS in the patient presented in the Grand Rounds article. In each study, the subgroup analysis was done using the factor of interest (age or sex), but these treatment-by-factor relationships were not affected by symptomatic status. Their assessment is mitigated by 3 factors: (1) lack of statistical power, (2) the treatment of stroke as an event of singular severity, and (3) the use of data that combines symptomatic and asymptomatic patients. First, Paraskevas and Veith report that symptomatic status does not affect the outcome; however, that would mistake proof for lack of statistical power. As noted in the substudy of CREST by symptomatic status,2 the total number of strokes among the 1181 asymptomatic participants was 23 (15 receiving CAS and 8 receiving CEA). The event rate was low and the difference was not significant, with the HRs ranging from 1.50 to 2.06. Any attempt to analyze additional subgroups, whether by age or sex, would only make the event total smaller, further limiting the statistical power to understand the effect of age or sex in asymptomatic patients. Second, Paraskevas and Veith describe stroke as though it is a singular item. In CREST, both minor and major stroke events are reported.2 Minor strokes, which accounted for 18 of the 23 total stroke events, were defined as those in which patients were asymptomatic by 30 days. Major stroke, which accounted for 5 total events in the 1181 patient cohort, was defined as a stroke with 9 or more points at 3 months on the National Institutes of Health Stroke Scale. CREST shows that
Corresponding Author: Joshua A. Beckman, MD, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115 ([email protected]).
1. Brott TG, Hobson RW II, Howard G, et al; CREST Investigators. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med. 2010;363(1):11-23. 2. Silver FL, Mackey A, Clark WM, et al; CREST Investigators. Safety of stenting and endarterectomy by symptomatic status in the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST). Stroke. 2011;42(3):675-680. 3. Ritter MA, Jurk K, Schriek C, et al. Microembolic signals on transcranial Doppler ultrasound are correlated with platelet activation markers, but not with platelet-leukocyte associates: a study in patients with acute stroke and in patients with asymptomatic carotid stenosis. Neurol Res. 2009;31(1):11-16. 4. Kinsella JA, Tobin WO, Cox D, et al. Prevalence of ex vivo high on-treatment platelet reactivity on antiplatelet therapy after transient ischemic attack or ischemic stroke on the PFA-100(®) and VerifyNow(®). J Stroke Cerebrovasc Dis. 2013;22(7):e84-e92. 5. Abbas A, Aukrust P, Dahl TB, et al. High levels of S100A12 are associated with recent plaque symptomatology in patients with carotid atherosclerosis [published correction appears in Stroke. 2013;44(7):e82]. Stroke. 2012;43(5):1347-1353. 6. Koutouzis M, Rallidis LS, Peros G, et al. Serum interleukin-6 is elevated in symptomatic carotid bifurcation disease. Acta Neurol Scand. 2009;119(2):119-125.
Diagnosis of Neonatal Infection With Herpes Simplex Virus To the Editor In the JAMA Clinical Challenge presenting an infant with neonatal infection with herpes simplex virus (HSV), Dr Ladizinski and colleagues1 recommended performing rapid direct fluorescence antibody (DFA) testing from a vesicle to establish the diagnosis. Ladizinski et al1 stated that DFA should be chosen over alternative diagnostic techniques, including viral cultures and polymerase chain reaction (PCR) assays, because DFA provides results most rapidly. However, recent recommendations from both the Infectious Diseases Society of America and the American Society for Microbiology characterize nucleic acid amplification test-
jama.com
JAMA February 5, 2014 Volume 311, Number 5
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of St. Andrews Library User on 06/07/2015
527
Letters
ing, such as PCR, as the “most sensitive, specific, and rapid test for diagnosis of HSV-associated skin or mucosal lesions.”2 These characteristics of PCR explain why our institution’s laboratory no longer offers in-house DFA or viral culture as techniques to diagnose cutaneous HSV. Obtaining a specimen for PCR is not difficult; one simply scrapes the base of a cutaneous or mucosal lesion using a swab.2 A PCR assay also has the ability to differentiate between HSV types 1 and 2.2 Additionally, the most recent guidelines from the US Centers for Disease Control and Prevention omit DFA entirely from their discussion on the evaluation of possible HSV infection.3 Instead, they recommend PCR or viral culture as the preferred tests, adding that the sensitivity of PCR is superior to that of culture.3 Thus PCR, rather than DFA, is the diagnostic test of choice. The physician’s decision regarding which diagnostic test to order depends, ultimately, on the laboratory capabilities of the institution, specifically which tests are available and how frequently the tests are performed. Few hospital laboratories run HSV diagnostic tests on a per case basis and most generally perform (or send out) the tests infrequently during the work week and not on weekends at all. Under these circumstances, consultation with the clinical laboratory can help determine which test will provide the answer soonest. All else being equal, it seems that PCR should be regarded as the procedure of choice to diagnose neonatal HSV. Antiviral therapy should be initiated while waiting for the test results in a case of suspected neonatal HSV. Nathan W. Rojek, BS Scott A. Norton, MD, MPH Author Affiliations: Georgetown University School of Medicine, Washington, DC (Rojek); Department of Dermatology, Children’s National Medical Center, Washington, DC (Norton). Corresponding Author: Nathan W. Rojek, BS, Georgetown University School of Medicine, 2456 Tunlaw Rd NW, Washington, DC 20007 ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Ladizinski B, Rukhman E, Lee KC. A 4-day-old neonate with a widespread vesicular rash. JAMA. 2013;310(9):971-972. 2. Baron EJ, Miller JM, Weinstein MP, et al. A guide to utilization of the microbiology laboratory for diagnosis of infectious diseases: 2013 recommendations by the Infectious Diseases Society of America (IDSA) and the American Society for Microbiology (ASM)(a). Clin Infect Dis. 2013;57(4):e22e121. 3. Centers for Disease Control and Prevention. Diseases characterized by genital, anal, or perianal ulcers. http://www.cdc.gov/std/treatment/2010 /genital-ulcers.htm#hsv. Accessed September 6, 2013.
In Reply Even though rapid DFA testing is not the reference standard test for diagnosing HSV, it was the best option among those given in the answer choices, and the most rapid diagnostic test available because PCR was not an option at the hospital during the weekend when the infant presented.
528
Given that PCR is the preferred diagnostic method based on current recommendations,1,2 it was ordered during the week. Meanwhile, the rapid DFA test was sent due to the availability of the test. This case highlights the utility of additional testing methods for diagnosis of neonatal HSV. Barry Ladizinski, MD Kachiu C. Lee, MD, MPH Author Affiliations: Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland (Ladizinski); Brown University, Providence, Rhode Island (Lee). Corresponding Author: Barry Ladizinski, MD, Johns Hopkins Bloomberg School of Public Health, 300 N Charles St, Ste 304, Baltimore, MD 21201 ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Centers for Disease Control and Prevention. Diseases characterized by genital, anal, or perianal ulcers. http://www.cdc.gov/std/treatment/2010 /genital-ulcers.htm#hsv. Accessed October 7, 2013. 2. Baron EJ, Miller JM, Weinstein MP, et al. A guide to utilization of the microbiology laboratory for diagnosis of infectious diseases: 2013 recommendations by the Infectious Diseases Society of America (IDSA) and the American Society for Microbiology (ASM)(a). Clin Infect Dis. 2013;57(4):e22-e121.
CORRECTION Patient Data Switched in Table: In the Original Investigation titled “Risk of Major Adverse Cardiac Events Following Noncardiac Surgery in Patients With Coronary Stents,” published in the October 9, 2013, issue of JAMA (2013;310[14]:14621472. doi:10.1001/jama.2013.278787), the numbers of male and female patients were switched in the first column of data in Table 1. This article has been corrected online.
Guidelines for Letters Letters discussing a recent JAMA article should be submitted within 4 weeks of the article's publication in print. Letters received after 4 weeks will rarely be considered. Letters should not exceed 400 words of text and 5 references and may have no more than 3 authors. Letters reporting original research should not exceed 600 words of text and 6 references and may have no more than 5 authors. They may include up to 2 tables or figures but online supplementary material is not allowed. All letters should include a word count. Letters must not duplicate other material published or submitted for publication. Letters not meeting these specifications are generally not considered. Letters being considered for publication ordinarily will be sent to the authors of the JAMA article, who will be given the opportunity to reply. Letters will be published at the discretion of the editors and are subject to abridgement and editing. Further instructions can be found at http://jama.com/public /InstructionsForAuthors.aspx. A signed statement for authorship criteria and responsibility, financial disclosure, copyright transfer, and acknowledgment and the ICMJE Form for Disclosure of Potential Conflicts of Interest are required before publication. Letters should be submitted via the JAMA online submission and review system at http: //manuscripts.jama.com. For technical assistance, please contact [email protected]. Section Editor: Jody W. Zylke, MD, Senior Editor.
JAMA February 5, 2014 Volume 311, Number 5
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of St. Andrews Library User on 06/07/2015
jama.com
