News & Analysis Medical News & Perspectives......p1783
Capitol Health Call ........................p1787
News From the CDC......................p1788
Everything You Ever Wanted to Know About Evidence-based Medicine
Bill Takes Aim at Nationwide Shortage of Nurses
ED Visits for Strokes Decline While Imaging Rates Increase
The JAMA Forum ..........................p1785
Senate Extends Funding That Provides Health Insurance for Children
Too Many Sidelined by COPD
Of SCOTUS and Chicken
Senators Introduce Legislation to Streamline Development and Approval of New Treatments GAO: The Way Medicare Pays Certain Hospitals Could Lead to Overspending
Medical News & Perspectives
Everything You Ever Wanted to Know About Evidence-based Medicine Rebecca Voelker, MSJ
McMaster University
W
hat’s in a name? Just ask Gordon Guyatt, MD, MSc. Twenty-five years ago, when he took over as residency director of the internal medicine program at McMaster University in Hamilton, Ontario, Canada, Guyatt wanted to teach resident physicians how to find the best clinical evidence in the medical literature, interpret it correctly, and practice it with patients. “It was something very new at the time,” he says. To spread the word and appeal to Canadian medical school graduates, he needed a simple label for the approach. He suggested the term “scientific medicine” during a Department of Medicine meeting. “The folks at the meeting, particularly the basic scientists, were outraged,” he recalls. They certainly weren’t teaching unscientific medicine, they protested, and the basic scientists already considered their work to be scientific medicine. So Guyatt needed a new name. Back at the drawing board, he came up with “evidence-based medicine.” He encountered no backlash and the name stuck. “It proved a remarkably good choice,” Guyatt says. The term was first published in the American College of Physicians’ Journal Club in 1991 and then in JAMA the following year. In 1993, an ongoing series called the Users’ Guides to the Medical Literature was launched in JAMA. The series gave rise in 2002 to the first edition of a book with the same name and an added descrip-
tor: A Manual for Evidence-based Clinical Practice. The second edition appeared in 2008, and earlier this year the third edition was published. Guyatt, distinguished professor of clinical epidemiology and biostatistics and of medicine at McMaster and principal editor of the books, recently spent time with JAMA to discuss updates in the newest edition. An edited version of the conversation follows. JAMA: Why was an update to the Users’ Guides needed at this time? DR GUYATT: It had been 6 years since the previous edition, and we had published a number of new Users’ Guides addressing current and emerging issues in the medical literature. We wanted to include these in a new book, and there were some chapters where real changes or advances had been made in the methodology that we wanted to incorporate as well. JAMA: Could you describe those advances? DR GUYATT: One is noninferiority trials. In traditional clinical trials, you have a new treatment that you think reduces deaths or strokes or heart attacks or improves quality of life. More recently, new treatments have been introduced not because they have good effects on those primary outcomes, but rather because they have other benefits: reducing side effects, being more convenient, or reducing bur-
jama.com
Gordon Guyatt, MD, MSc
den. For example, novel anticoagulants were introduced not because they reduce strokes or other thrombotic events, but because they don’t have the monitoring and drug-interaction problems that you see with warfarin. With these new treatments comes a new design for clinical trials. We try not to show that the new drugs are better, but that they’re not too much worse. These have been called noninferiority trials, and they raise new challenges in interpreting and understanding treatment for the condition. Quality improvement also is a big issue now in medicine. Trials are appearing about quality improvement or quality assurance, procedures, and initiatives. So
(Reprinted) JAMA May 12, 2015 Volume 313, Number 18
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Carleton University User on 05/13/2015
1783
News & Analysis
we have a chapter that deals with the critical appraisal and the use of studies about quality improvement. Another emerging area of which there are now many studies are genetic association studies examining whether particular genes are associated with bad outcomes or with the development of conditions, like genes that may be associated with the development of Alzheimer disease. We have a chapter in the new book on these kinds of studies. Other new chapters address additional important advances: one chapter covers shared decision making between patients and clinicians, which more and more is becoming an important part of evidence-based clinical practice. And there are 2 chapters on systematic reviews and meta-analyses. The first of these chapters defines the difference between a systematic review and a meta-analysis and explains how to judge the trustworthiness of the systematic review meta-analytic processes. As we note, a systematic review is a
choose? In the last decade, a new statistical approach, the network meta-analysis, has been developed. It allows the simultaneous comparison of A vs B vs C vs D vs E vs F, and so on. In the new book, we explain the concept of network metaanalysis and how the clinician can interpret these analyses. Many, many have been published very quickly just in the last several years. Clinicians are going to see them more and more, so they need a Users’ Guide.
JAMA: Which chapter do you think clinicians will find most useful when they’re looking for answers? DR GUYATT: An important change in the existing chapters is on finding the evidence— searching for the best evidence to answer your question. Historically we sent people to Medline for individual articles, but what we suggest now is to look first for preprocessed evidence in advanced medical texts, summaries, practice guidelines, and online clinical reference tools like DynaMed, CliniWe hope that even if clinicians are not cal Evidence, Innecessarily aware of evidence-based foRetriever, and PDxMD. We are principles and how to use them to now providing assess the literature themselves, they guidance for how to quickly will still use evidence-based guidelines and efficiently that lead to evidence-based practice. access and interpret high-qualsummary of research from different studies ity preprocessed evidence. If these sources that addresses a focused clinical question in fail, you can still go on to more traditional rea systematic, reproducible manner, and this sources like a Medline search. Another tool may be accompanied by a meta-analysis, that has revolutionized ways to find eviwhich is a statistical pooling or aggregation dence is updating services that send noof results from different studies to provide tices of new and exciting things happening a single best estimate of effect of a therapy, in your field. Many are free and they proprognosis, or diagnostic test. The second vide regular updates on what is new in your chapter explains how to trust and apply the area of practice. results of systematic reviews and metaanalyses. Those are some of the major JAMA: President Obama recently anchanges that really enhance the value of nounced a new precision medicine initiathe Users’ Guides. tive to study how genes, health, and the environment are linked. What role can JAMA: The new edition also includes a evidence-based medicine play in this initiachapter on network meta-analysis. Why is tive? that important? DR GUYATT: What evidence-based mediDR GUYATT: Typically, treatments for a cine will do is help people exercise the particular condition have been compared appropriate caution in using the results of to placebo or 1 drug to another. But there ongoing studies. Just because a gene is may be 6 or even a dozen drugs available associated with an outcome does not for that condition. How does the clinician mean that everybody needs to be tested 1784
for it or that the information is going to be valuable. There’s a lot of exciting potential around genetic testing and using the information in terms of guiding our diagnostic and therapeutic approaches. But there’s also a lot of potential for wasted resources and even for harm. So the relation with evidence-based medicine is that if you understand the principles of evidencebased medicine, you are going to be able to distinguish between new tests or procedure s or diagnostic strategie s that improve patient outcomes and those that simply use up resources with no benefit and possible harm. JAMA: How well do you think US medical schools and those in other countries are doing in teaching evidence-based medicine? DR GUYATT: Better and better, but still a long way to go. For the last 4 years, I have been chairing a panel that is bringing evidence-based medicine questions to the United States Medical Licensing Examination. We are using an innovative approach. Instead of asking knowledge questions, we present abstracts like those in the American College of Physicians’ Journal Club—good, preprocessed evidence. Then we ask questions about the test takers’ interpretation to see if they understand it. The results are not always terribly encouraging, so we try to keep our questions reasonably simple and straightforward. But from the results we are getting, we have to make them even simpler and more straightforward. This tells us that the concepts of evidence-based medicine for many medical schools either are not a high priority or are not being taught optimally. Sometimes what is being tested drives curriculum, so the fact that these are now appearing on the exam may heighten the profile of teaching evidence-based medicine concepts in US medical schools. JAMA: Along similar lines, can you gauge how well physicians are practicing evidencebased medicine? DR GUYATT: There are very few studies that really tell us about that, so my responses are based only on anecdotes and very lowquality evidence. Having said that, I think it’s better in younger clinicians who have at least had some training in this area and are perhaps more vividly aware of its importance. We hope that even if clinicians are not nec-
JAMA May 12, 2015 Volume 313, Number 18 (Reprinted)
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Carleton University User on 05/13/2015
jama.com
News & Analysis
essarily aware of evidence-based principles and how to use them to assess the literature themselves, they will still use evidence-based guidelines that lead to evidence-based practice. JAMA: Evidence-based medicine once was criticized as “cookbook medicine.” Do you still encounter that kind of criticism, and if so, how do you counter it?
DR GUYATT: One of my favorite talks to give is that evidence-based medicine is patient-centered medicine, so there are a number of things to point out. Perhaps most important is that evidence itself never tells you what to do, never. It’s always evidence in the context of values and preferences. In other words, a person with one set of values and preferences will say, “Yes, it’s the right thing.” Another
informed individual would say, “No. For me, the undesirable consequences outweigh the desirable consequences.” Evidence-based medicine has highlighted the fact that so often decisions are value and preference sensitive. To do the best for the individual patient, you need to take into account their values and preferences, ideally in the context of shared decision making.
The JAMA Forum
Of SCOTUS and Chicken Larry Levitt, MPP
T
he game of chicken, which was popularized in the 1950s movie Rebel Without a Cause, has many variants, but the basic design goes like this: players involved in a conflict of some sort try not to yield in the hope that the other player will yield first. But the worst and potentially catastrophic outcome is when no one yields. After hearing oral arguments on March 4, the US Supreme Court (aka SCOTUS) is deliberating in King v Burwell, a case that has the potential to unleash a massive game of chicken around the Affordable Care Act (ACA). The case centers (http://bit.ly /1FEWZu4) on circumstances related to premium subsidies under the ACA, which are now available to people with low and moderate incomes in all states. King v Burwell challenges the legality of subsidies in states where the federal government set up an exchange because the state declined to set up its own health insurance marketplace.
American Medical Association
Swift, Severe Consequences The consequences (http://bit.ly/1ERCica) of a decision in favor of the challengers would be swift and severe: • Subsidies would end—likely within a month of a decision, which is expected in late June—for about 7.5 million people who now qualify for them in the 34 states not running their own marketplaces (in the 17 state-based marketplaces, nothing would change and subsidies would continue). • The premiums these 7.5 million people pay for insurance, would rise from an average
of $105 (after taking the subsidies into account) to $374 per month, an increase of 256%. • People who are sick and know they need insurance would likely work hard to find a way to keep it, but those who are healthy would likely drop it. The ACA’s individual mandate—which is the stick to get healthy people to enroll, working hand-in-hand with the carrot of the subsidies—would be largely ineffective. That’s because 83% of uninsured individuals who are currently eligible for subsidies would be exempt from the requirement to have coverage because it would be unaffordable without the subsidies. • The result in affected states would be a classic “death spiral.” Premiums would rise, more healthy people would drop their coverage, and that in turn would cause premiums to rise even more. This would destabilize the whole individual market in these states because insurers are required to set premiums within a state based on their entire individual market business, not just people buying through the marketplace. No one will want to yield in this scenario. But the consequences of no one yielding are indeed dire. The important thing to understand about the King v Burwell case is that it does not (at least as it’s been argued before the Court) involve the constitutionality of the ACA. Rather, it’s a matter of statutory interpretation: did the Internal Revenue Service have the authority under the law to provide subsidies in all states?
jama.com
Larry Levitt, MPP
That means with just a few strokes on the keyboard, Congress could clarify that subsidies should be provided to people in state-based and federal marketplaces alike. Such a swerve would avoid a catastrophe quickly and easily. But with many Republicans in Congress adamantly opposed to Obamacare, no one expects such a yield. Enter the 34 governors and state legislatures that have not set up their own marketplaces under the ACA. If they were to yield and create state-based marketplaces, they would render moot a possible Supreme Court ruling against subsidies in states without their own exchange. To be sure, there would be strong pressure on states to take this route. Many of their residents would lose insurance if they
(Reprinted) JAMA May 12, 2015 Volume 313, Number 18
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Carleton University User on 05/13/2015
1785
Capitol Health Call ........................p1787
News From the CDC......................p1788
Everything You Ever Wanted to Know About Evidence-based Medicine
Bill Takes Aim at Nationwide Shortage of Nurses
ED Visits for Strokes Decline While Imaging Rates Increase
The JAMA Forum ..........................p1785
Senate Extends Funding That Provides Health Insurance for Children
Too Many Sidelined by COPD
Of SCOTUS and Chicken
Senators Introduce Legislation to Streamline Development and Approval of New Treatments GAO: The Way Medicare Pays Certain Hospitals Could Lead to Overspending
Medical News & Perspectives
Everything You Ever Wanted to Know About Evidence-based Medicine Rebecca Voelker, MSJ
McMaster University
W
hat’s in a name? Just ask Gordon Guyatt, MD, MSc. Twenty-five years ago, when he took over as residency director of the internal medicine program at McMaster University in Hamilton, Ontario, Canada, Guyatt wanted to teach resident physicians how to find the best clinical evidence in the medical literature, interpret it correctly, and practice it with patients. “It was something very new at the time,” he says. To spread the word and appeal to Canadian medical school graduates, he needed a simple label for the approach. He suggested the term “scientific medicine” during a Department of Medicine meeting. “The folks at the meeting, particularly the basic scientists, were outraged,” he recalls. They certainly weren’t teaching unscientific medicine, they protested, and the basic scientists already considered their work to be scientific medicine. So Guyatt needed a new name. Back at the drawing board, he came up with “evidence-based medicine.” He encountered no backlash and the name stuck. “It proved a remarkably good choice,” Guyatt says. The term was first published in the American College of Physicians’ Journal Club in 1991 and then in JAMA the following year. In 1993, an ongoing series called the Users’ Guides to the Medical Literature was launched in JAMA. The series gave rise in 2002 to the first edition of a book with the same name and an added descrip-
tor: A Manual for Evidence-based Clinical Practice. The second edition appeared in 2008, and earlier this year the third edition was published. Guyatt, distinguished professor of clinical epidemiology and biostatistics and of medicine at McMaster and principal editor of the books, recently spent time with JAMA to discuss updates in the newest edition. An edited version of the conversation follows. JAMA: Why was an update to the Users’ Guides needed at this time? DR GUYATT: It had been 6 years since the previous edition, and we had published a number of new Users’ Guides addressing current and emerging issues in the medical literature. We wanted to include these in a new book, and there were some chapters where real changes or advances had been made in the methodology that we wanted to incorporate as well. JAMA: Could you describe those advances? DR GUYATT: One is noninferiority trials. In traditional clinical trials, you have a new treatment that you think reduces deaths or strokes or heart attacks or improves quality of life. More recently, new treatments have been introduced not because they have good effects on those primary outcomes, but rather because they have other benefits: reducing side effects, being more convenient, or reducing bur-
jama.com
Gordon Guyatt, MD, MSc
den. For example, novel anticoagulants were introduced not because they reduce strokes or other thrombotic events, but because they don’t have the monitoring and drug-interaction problems that you see with warfarin. With these new treatments comes a new design for clinical trials. We try not to show that the new drugs are better, but that they’re not too much worse. These have been called noninferiority trials, and they raise new challenges in interpreting and understanding treatment for the condition. Quality improvement also is a big issue now in medicine. Trials are appearing about quality improvement or quality assurance, procedures, and initiatives. So
(Reprinted) JAMA May 12, 2015 Volume 313, Number 18
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Carleton University User on 05/13/2015
1783
News & Analysis
we have a chapter that deals with the critical appraisal and the use of studies about quality improvement. Another emerging area of which there are now many studies are genetic association studies examining whether particular genes are associated with bad outcomes or with the development of conditions, like genes that may be associated with the development of Alzheimer disease. We have a chapter in the new book on these kinds of studies. Other new chapters address additional important advances: one chapter covers shared decision making between patients and clinicians, which more and more is becoming an important part of evidence-based clinical practice. And there are 2 chapters on systematic reviews and meta-analyses. The first of these chapters defines the difference between a systematic review and a meta-analysis and explains how to judge the trustworthiness of the systematic review meta-analytic processes. As we note, a systematic review is a
choose? In the last decade, a new statistical approach, the network meta-analysis, has been developed. It allows the simultaneous comparison of A vs B vs C vs D vs E vs F, and so on. In the new book, we explain the concept of network metaanalysis and how the clinician can interpret these analyses. Many, many have been published very quickly just in the last several years. Clinicians are going to see them more and more, so they need a Users’ Guide.
JAMA: Which chapter do you think clinicians will find most useful when they’re looking for answers? DR GUYATT: An important change in the existing chapters is on finding the evidence— searching for the best evidence to answer your question. Historically we sent people to Medline for individual articles, but what we suggest now is to look first for preprocessed evidence in advanced medical texts, summaries, practice guidelines, and online clinical reference tools like DynaMed, CliniWe hope that even if clinicians are not cal Evidence, Innecessarily aware of evidence-based foRetriever, and PDxMD. We are principles and how to use them to now providing assess the literature themselves, they guidance for how to quickly will still use evidence-based guidelines and efficiently that lead to evidence-based practice. access and interpret high-qualsummary of research from different studies ity preprocessed evidence. If these sources that addresses a focused clinical question in fail, you can still go on to more traditional rea systematic, reproducible manner, and this sources like a Medline search. Another tool may be accompanied by a meta-analysis, that has revolutionized ways to find eviwhich is a statistical pooling or aggregation dence is updating services that send noof results from different studies to provide tices of new and exciting things happening a single best estimate of effect of a therapy, in your field. Many are free and they proprognosis, or diagnostic test. The second vide regular updates on what is new in your chapter explains how to trust and apply the area of practice. results of systematic reviews and metaanalyses. Those are some of the major JAMA: President Obama recently anchanges that really enhance the value of nounced a new precision medicine initiathe Users’ Guides. tive to study how genes, health, and the environment are linked. What role can JAMA: The new edition also includes a evidence-based medicine play in this initiachapter on network meta-analysis. Why is tive? that important? DR GUYATT: What evidence-based mediDR GUYATT: Typically, treatments for a cine will do is help people exercise the particular condition have been compared appropriate caution in using the results of to placebo or 1 drug to another. But there ongoing studies. Just because a gene is may be 6 or even a dozen drugs available associated with an outcome does not for that condition. How does the clinician mean that everybody needs to be tested 1784
for it or that the information is going to be valuable. There’s a lot of exciting potential around genetic testing and using the information in terms of guiding our diagnostic and therapeutic approaches. But there’s also a lot of potential for wasted resources and even for harm. So the relation with evidence-based medicine is that if you understand the principles of evidencebased medicine, you are going to be able to distinguish between new tests or procedure s or diagnostic strategie s that improve patient outcomes and those that simply use up resources with no benefit and possible harm. JAMA: How well do you think US medical schools and those in other countries are doing in teaching evidence-based medicine? DR GUYATT: Better and better, but still a long way to go. For the last 4 years, I have been chairing a panel that is bringing evidence-based medicine questions to the United States Medical Licensing Examination. We are using an innovative approach. Instead of asking knowledge questions, we present abstracts like those in the American College of Physicians’ Journal Club—good, preprocessed evidence. Then we ask questions about the test takers’ interpretation to see if they understand it. The results are not always terribly encouraging, so we try to keep our questions reasonably simple and straightforward. But from the results we are getting, we have to make them even simpler and more straightforward. This tells us that the concepts of evidence-based medicine for many medical schools either are not a high priority or are not being taught optimally. Sometimes what is being tested drives curriculum, so the fact that these are now appearing on the exam may heighten the profile of teaching evidence-based medicine concepts in US medical schools. JAMA: Along similar lines, can you gauge how well physicians are practicing evidencebased medicine? DR GUYATT: There are very few studies that really tell us about that, so my responses are based only on anecdotes and very lowquality evidence. Having said that, I think it’s better in younger clinicians who have at least had some training in this area and are perhaps more vividly aware of its importance. We hope that even if clinicians are not nec-
JAMA May 12, 2015 Volume 313, Number 18 (Reprinted)
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Carleton University User on 05/13/2015
jama.com
News & Analysis
essarily aware of evidence-based principles and how to use them to assess the literature themselves, they will still use evidence-based guidelines that lead to evidence-based practice. JAMA: Evidence-based medicine once was criticized as “cookbook medicine.” Do you still encounter that kind of criticism, and if so, how do you counter it?
DR GUYATT: One of my favorite talks to give is that evidence-based medicine is patient-centered medicine, so there are a number of things to point out. Perhaps most important is that evidence itself never tells you what to do, never. It’s always evidence in the context of values and preferences. In other words, a person with one set of values and preferences will say, “Yes, it’s the right thing.” Another
informed individual would say, “No. For me, the undesirable consequences outweigh the desirable consequences.” Evidence-based medicine has highlighted the fact that so often decisions are value and preference sensitive. To do the best for the individual patient, you need to take into account their values and preferences, ideally in the context of shared decision making.
The JAMA Forum
Of SCOTUS and Chicken Larry Levitt, MPP
T
he game of chicken, which was popularized in the 1950s movie Rebel Without a Cause, has many variants, but the basic design goes like this: players involved in a conflict of some sort try not to yield in the hope that the other player will yield first. But the worst and potentially catastrophic outcome is when no one yields. After hearing oral arguments on March 4, the US Supreme Court (aka SCOTUS) is deliberating in King v Burwell, a case that has the potential to unleash a massive game of chicken around the Affordable Care Act (ACA). The case centers (http://bit.ly /1FEWZu4) on circumstances related to premium subsidies under the ACA, which are now available to people with low and moderate incomes in all states. King v Burwell challenges the legality of subsidies in states where the federal government set up an exchange because the state declined to set up its own health insurance marketplace.
American Medical Association
Swift, Severe Consequences The consequences (http://bit.ly/1ERCica) of a decision in favor of the challengers would be swift and severe: • Subsidies would end—likely within a month of a decision, which is expected in late June—for about 7.5 million people who now qualify for them in the 34 states not running their own marketplaces (in the 17 state-based marketplaces, nothing would change and subsidies would continue). • The premiums these 7.5 million people pay for insurance, would rise from an average
of $105 (after taking the subsidies into account) to $374 per month, an increase of 256%. • People who are sick and know they need insurance would likely work hard to find a way to keep it, but those who are healthy would likely drop it. The ACA’s individual mandate—which is the stick to get healthy people to enroll, working hand-in-hand with the carrot of the subsidies—would be largely ineffective. That’s because 83% of uninsured individuals who are currently eligible for subsidies would be exempt from the requirement to have coverage because it would be unaffordable without the subsidies. • The result in affected states would be a classic “death spiral.” Premiums would rise, more healthy people would drop their coverage, and that in turn would cause premiums to rise even more. This would destabilize the whole individual market in these states because insurers are required to set premiums within a state based on their entire individual market business, not just people buying through the marketplace. No one will want to yield in this scenario. But the consequences of no one yielding are indeed dire. The important thing to understand about the King v Burwell case is that it does not (at least as it’s been argued before the Court) involve the constitutionality of the ACA. Rather, it’s a matter of statutory interpretation: did the Internal Revenue Service have the authority under the law to provide subsidies in all states?
jama.com
Larry Levitt, MPP
That means with just a few strokes on the keyboard, Congress could clarify that subsidies should be provided to people in state-based and federal marketplaces alike. Such a swerve would avoid a catastrophe quickly and easily. But with many Republicans in Congress adamantly opposed to Obamacare, no one expects such a yield. Enter the 34 governors and state legislatures that have not set up their own marketplaces under the ACA. If they were to yield and create state-based marketplaces, they would render moot a possible Supreme Court ruling against subsidies in states without their own exchange. To be sure, there would be strong pressure on states to take this route. Many of their residents would lose insurance if they
(Reprinted) JAMA May 12, 2015 Volume 313, Number 18
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Carleton University User on 05/13/2015
1785
