Diagnostic time out

What is a diagnostic time out? Succinctly defined, it’s a deliberate exercise in differential diagnosis and systematic clinical reasoning in the care of an individual patient. But wait, I hear someone say… isn’t that what we do already? Well, no. We’re all familiar with the traditional model for clinical reasoning that we’re taught in medical school but those of us in the real world of practice nowadays, if we’re honest, realize that it seldom happens. There’s just not enough time when you’re forced to see too many patients each day. And hospitalist incentives, with their emphasis on speed and quick adoption of specific diagnostic labels, run in opposition. What do we as hospitalists do instead? Well, aside from all the care pathways and metric incentives that tell us what to do, we rely on clinical instincts and rules of thumb. Because they bypass formal analysis, they save time. They serve as cognitive shortcuts. We call these heuristics. This method of thinking (fast, instinctive, intuitive) is sometimes known as system 1 thinking. It has the advantages of being efficient and fast and sometimes, in critical situations, life saving. But it comes at the cost of a certain error rate. In order to better understand the process of system 1 thinking we have given the various heuristics names and categories. I recently listed some of those in this post. 

If system 1 is our usual measure of processing to get around time constraints the alternative is system 2: formal clinical reasoning .  System 2 thinking was the topic of a recent paper in CriticalCare Clinics. Although based on a survey of people working in a NICU the article has general applicability. The authors contrast system 1 and system 2 thinking in this manner:

Dual process theory holds that individuals engaging in medical decision-making use one of 2 distinct cognitive processes: a system 1 process based on heuristics – the use of rapid pattern recognition and rules of thumb – or a system 2 process, based on deliberate analytical modeling and hypothesis generation. While invoking system one processes individuals can think fast and reflexively and can even operate at a subconscious level, using pattern recognition to sort vast amounts of clinical information quickly before an illness script that allows for the rapid elaboration of a differential diagnosis. In contrast system 2 processes require focused attention and are purposefully analytical, relying on deliberate counter-factual reasoning to generate hypotheses regarding the pathophysiologic mechanisms by which a patient’s symptoms are produced.

The authors introduced the concept of the diagnostic time out to describe this shift of thinking because it requires deliberate effort. It’s not going to arise spontaneously in the natural course of the ward routine. (The authors were not the first ones to use this term). The diagnostic time out can be considered the cognitive equivalent of the better known procedural time out.

Why is a diagnostic time out needed? Research on diagnostic error has indicated that while some instances are due to system problems (such as failure to communicate test results) most are cognitive errors. These can be linked to the heuristics of system 1 thinking. The diagnostic time out, or the deliberate exercise of system 2 thinking, is a way to complement these cognitive shortcuts with a more analytical process.

Some opinion leaders in the field of diagnostic error have suggested universal adoption of system 2 thinking. This is problematic due to time constraints. Besides, there are some essential benefits of system 1 thinking, particularly in acute life-threatening situations. The real trick is how best to selectively employ system 2 thinking. In other words what are the situations in which system 2 thinking should be used? The authors suggest handoff situations in complex patients including ER to hospitalist, off service/on service and ICU to ward transfers.

How does it work? The authors propose a template but it’s really just the traditional clinical reasoning process. One of their points really got my attention: during the time out diagnostic labels should be removed and replaced by signs, symptoms, manifestations and clinical concerns. This of course is the opposite of what your coders and hospitalist leaders want you to do.

What are some of the barriers to implementation? In addition to time constraints, fear of ambiguity is an important factor. We are afraid to admit what we don’t know. One thing you will never hear a hospitalist say out loud is “I’ll have to think about that.”

A little more on metacognition

This article from Academic Emergency Medicine, published in 2002, remains applicable today. It makes the point that heuristics in medicine are valuable even though they can lead to error. The article also makes the statement:

The increasing use of clinical decision rules, as well as other aids that reduce uncertainty and cognitive load, e.g., computerized clinical decision support,will improve certain aspects of clinical decision making, but much flesh-and-blood clinical decision making will remain and there will always be a place for intuition and clinical acumen.

It presents an exhaustive list with detailed descriptions of the various cognitive shortcuts.

Indulge me in a little metacognition

I found an interesting post about cognitive shortcuts in medicine. I have a minor objection to the title of the post which is Cognitive Errors. Cognitive shortcuts, known as heuristics, which are examples of fast instinctive thinking (system one), often lead to error. In some cases, however, they can be useful because they are efficient and time saving. There is an up side as well as a down side to system one thinking in medicine.

Let’s go down the list. I’ve skipped some of them.

The first example given is affective error. This refers to an emotional response surpassing objectivity.

Next is aggregate bias. I struggle with this one. The author says that the aggregate bias is the belief that data in the aggregate don’t apply to the patient in front of you. My understanding (maybe I’m wrong) is that aggregate bias, otherwise known as the ecological fallacy, is the opposite. That is, it refers to inappropriate application of population data to an individual. It has more to do with treatment decisions than diagnostic error. Remember, one of the first principles of evidence-based medicine is that clinical reasoning decision making starts with the unique aspects of the individual patient. After looking at a variety of references, it would appear that both definitions have been used. Most medical references define aggregation bias the way the blog author does. Those outside of medicine define it as inappropriate extrapolation.

The ambiguity effect is really a bias against ambiguity. So we tend to stick with things we are more familiar with. That may cause us to ignore other possibilities and take too narrow a view of things. As originally conceived it had to do with probability. That is, people have a tendency to gravitate toward choices in which the probability is known or explicitly stated. Of note, the ambiguity effect was first described by Daniel Ellsberg.

The anchoring heuristic is one of the better known cognitive biases. This refers to the tendency to stick with one’s initial hunch despite new evidence to the contrary. You may be so proud of your initial hunch that you ignore new information. Confirmation bias and diagnostic momentum are related concepts.

Ascertainment bias, as the author points out, is an umbrella category. It encompasses a lot of stereotypes and biases. In essence it’s just—-well, bias. It’s not very useful as a unique category in discussions of cognitive error.

Availability bias is one of the better known cognitive shortcuts. This refers to the influence of prior experience. This causes bias toward the first thing that comes to your mind. For example, if you’ve been burned by having missed a case of aortic dissection you may tend to be over concerned about aortic dissection in every future case of chest pain. The flip side is you may fail to consider things you haven’t seen in a long time.

Base rate neglect is a cognitive shortcut that may be considered harmful and wasteful in ambulatory medicine but may be your friend in the arena of hospital and emergency medicine. It’s a failing to consider the true prevalence of diseases in clinical reasoning. It ignores the old aphorism “common things happen most often.” In the high acuity world of the hospital, where you really need to be risk-averse, base rate neglect may be beneficial. Put another way you and and your patient may be better off if you consider worst case scenario.

Then there’s belief bias. I’m not sure this belongs in a discussion of diagnostic shortcuts as it has more to do with treatment recommendations. I cringe when I hear somebody say they “believe“ in a particular treatment, implying that belief surpasses reasoning from evidence .

Blind spot bias is similar to the Dunning Kruger effect in which we think we're smarter than we really are. Humility is the remedy here. Does this lead to a form of cognitive shortcut? Maybe in that we fail to pause and consider carefully that we might be wrong.

Confirmation bias is akin to anchoring. This is the tendency to be selective in what type of accumulating evidence you consider. That is, you consider mainly evidence that supports your original hunch.

The framing heuristic is another well known shortcut. We are biased toward diagnostic possibilities in accordance with the way the initial presentation is framed. Though it can be useful it restricts our differential diagnosis in a way that excludes a wide range of possibilities. Not every returning travel with fever has a parasite, for example.

The gamblers fallacy, according to the blog author, is “the erroneous belief that chance his self correcting.“ This is a cognitive error that tends in the opposite direction to the availability heuristic.

The order effect is something I was vaguely aware of but had not considered as a cognitive error category. It refers to the tendency to focus on information that is proximate in time and to do so at the expense of the totality of events over time. This typically occurs at the point of hand off in a patient who has had a very long hospital course.

Premature closure is just what it says. It’s a tendency for thinking to stop once a tentative diagnosis has been made. It overlaps with other categories such as anchoring. There is probably a subtle difference between premature closure and anchoring. Anchoring implies an emotional attachment to a diagnosis whereas premature closure implies diagnostic laziness.

Representativeness restraint has also been known as a representativeness heuristic. It is a cognitive shortcut characterized by focusing too much on the prototypical manifestations of a disease. This may cause the clinician to miss atypical presentations.

Search satisfaction is another example of laziness in clinical reasoning. It’s a tendency to stop searching once an answer has been found. The author gives the example of missing a second fracture on an x-ray once the first one is identified.

Sunk cost fallacy is a type of emotional heuristic as well as diagnostic laziness. It is the tendency to ignore new information and not consider alternative diagnoses once the original diagnosis has been arrived at after a great time effort and expense (the sunk cost).

Sutton’s slip might be the dark side of Sutton’s law (going where the money is). Pursuing the obvious might lead to error because of other possibilities being ignored.

Zebra retreat is the avoidance of rare diagnoses to a fault. It’s an opposite of base rate neglect.

Atul Gawande on the electronic medical record

Atul Gawande has a piece in the New Yorker titled Why Doctors Hate their Computers. The title is deceptive. In the first place doctors don’t hate computers (I’ve never met one who did, have you?). In the body of the paper Gawande doesn’t even seem to attempt to make that case. He does point out how doctors hated the way in which they were forced to adopt health information technology and the culture that went alongside. But, though he talks around it (and he talks a lot around it) he fails to answer the question of why. Is there something wrong with computers themselves in the current state of development? Is it the way policymakers and administrators have forced the implementation? Or is it that docs just need an attitude adjustment? He implies a little of each. Overall the article is incoherent.

Gawande has thrown together a mishmash of anecdotes, unreferenced claims and quotes from supposed experts. And the qualifications of these experts? Well, consider this one:

Gregg Meyer sympathizes, but he isn’t sorry. As the chief clinical officer at Partners HealthCare, Meyer supervised the software upgrade. An internist in his fifties, he has the commanding air, upright posture, and crewcut one might expect from a man who spent half his career as a military officer.

Hmmm. A commanding air, an upright posture and a crewcut. I think I’m afraid of this guy. He seems to think doctors have too much autonomy and a bad attitude to boot. He says:

“But we think of this as a system for us and it’s not,” he said. “It is for the patients.” 

Emphasis his.

Meyer just gave himself away. He’s operating on the idea that the interests of doctors are opposed to the interests of patients. It’s an ethical question worth pondering but not a great starting premise. Gawande seems to accept it uncritically. A little further on Gawande says of Meyer, also uncritically:

Gregg Meyer is understandably delighted to have the electronic levers to influence the tens of thousands of clinicians under his purview. He had spent much of his career seeing his hospitals blighted by unsafe practices that, in the paper-based world, he could do little about.

Evidence based medicine, particularly its third pillar (the importance of the expertise of the individual clinician) opposes such a top down approach. Does Gawande see anything wrong with Meyer’s line of thinking? If he does he doesn’t say so.

It’s style over substance:

Jessica Jacobs, a longtime office assistant in my practice—mid-forties, dedicated, with a smoker’s raspy voice—

As if that’s supposed to be a convincer in some way. But what does it mean, exactly? That she’s got savvy? That her dedication to her work has taken its toll? It’s left to our imagination.

Gawande fails to even come close to making the case that doctors hate computers, let alone answer the question of why, but he does point out some of the negative consequences of the EMR. Maybe this is progress, because it would have been nearly forbidden speech about a decade ago.

Annals article on EBM and machine learning makes huge category errors

From the beginning of the paper:

Machine learning (ML), which converts complex data into algorithms, challenges the traditional epidemiologic approach of evidence-based medicine (EBM). Here I outline the differences, strengths, and limitations of these 2 approaches and suggest areas of reconciliation.

Beginning in the 1970s, scientists extolled the virtues of EBM's hypothesis-driven, protocolized experiments involving well-defined populations and preselected exposure and outcome variables. Inferences were made using traditional biostatistics.

First, EBM wasn’t around in the 70’s.  It started, as a movement, in 1992.  Second, EBM is not a method of research.   It is, rather, a system of approaching the clinical care of individual patients.  One patient at a time.  The rest of the article is equally confusing.

Limitations of clinical trials with surrogate endpoints

Recently published:

Abstract

Background Surrogate endpoint trials test strategies more efficiently but are accompanied by uncertainty about the relationship between changes in surrogate markers and clinical outcomes.

Methods and Results We identified cardiovascular trials with primary surrogate endpoints published in the New England Journal of Medicine, Lancet, and JAMA: Journal of the American Medical Association from 1990 to 2011 and determined the trends in publication of surrogate endpoint trials and the success of the trials in meeting their primary endpoints. We tracked for publication of clinical outcome trials on the interventions tested in surrogate trials. We screened 3016 articles and identified 220 surrogate endpoint trials. From the total of 220 surrogate trials, 157 (71.4%) were positive for their primary endpoint. Only 59 (26.8%) surrogate trials had a subsequent clinical outcomes trial. Among these 59 trials, 24 outcomes trial results validated the positive surrogates, whereas 20 subsequent outcome trials were negative following positive results on a surrogate. We identified only 3 examples in which the surrogate trial was negative but a subsequent outcomes trial was conducted and showed benefit. Findings were consistent in a sample cohort of 383 screened articles inclusive of 37 surrogate endpoint trials from 6 other high‐impact journals.

Conclusions Although cardiovascular surrogate outcomes trials frequently show superiority of the tested intervention, they are infrequently followed by a prominent outcomes trial. When there was a high‐profile clinical outcomes study, nearly half of the positive surrogate trials were not validated. Cardiovascular surrogate outcome trials may be more appropriate for excluding benefit from the patient perspective than for identifying it.

The campaign against reperfusion therapy for stroke

It’s the most massive push-back I’ve ever seen in my career. It’s profiled in this NYT article. This isn’t new, and I’m surprised the popular media hasn’t caught hold of it before now. The opinion leader is Dr. Jerome Hoffman, professor of emergency medicine at UCLA. The article has drawn over 200 comments so far. I thought this one was interesting from an ER doc:

As an ER Doc, I am disheartened by the continued resistance by some in my specialty who to chose to deny this treatment option for appropriate patients. There are risks and benefits to every medical intervention but in this case many of us overstate the former and underestimate the latter.

I have personally debated with Dr. Hoffman (yes, he is charismatic but no, he most certainly did not win over that audience) and despite his assertions about case selection in the original trial, additional trials over the ensuring 22 years since FDA approval have repeatedly confirmed tPA as the international standard of care for acute ischemic stroke.

The concerns about the 11 "negative trials" are ludicrous. Much was unknown at the time these were conducted: they looked at higher doses, different drugs, and longer time eligibility windows that, if used today, would be malpractice. Yet these 11 studies were critical in defining today's standard of care. Therefore it is extremely disingenuous to cite them as evidence to justify denying our patients this critical intervention today.

Intention to treat analysis in RCTs

What is it? Why is it important?

Are we practicing EBM in the treatment of ARDS?

Far from it, it seems. Just over a quarter of patients with ARDS in the study got low tidal volume ventilation. Moreover, only about half the patients meeting ARDS criteria were acknowledged as such.

Can someone explain to me why I need AHRQ?

According to this Kevin MD post we'll all suffer if it becomes marginalized anymore than it already is. I've never used it. What am I missing?

Glossary of critical appraisal

This piece has some helpful information. However it’s a bit imprecise, offering descriptions that “talk around” the various terms rather than actual definitions.

Perspective and common sense in managing type 2 diabetes in older individuals

A recent JAMA paper reviewed this topic. From the abstract:

Four large randomized clinical trials (RCTs), ranging in size from 1791 to 11 440 patients, provide the majority of the evidence used to guide diabetes therapy. Most RCTs of intensive vs standard glycemic control excluded adults older than 80 years, used surrogate end points to evaluate microvascular outcomes and provided limited data on which subgroups are most likely to benefit or be harmed by specific therapies. Available data from randomized clinical trials suggest that intensive glycemic control does not reduce major macrovascular events in older adults for at least 10 years. Furthermore, intensive glycemic control does not lead to improved patient-centered microvascular outcomes for at least 8 years.

The surrogate endpoints for microvascular disease are things like findings of retinopathy or laboratory evidence of renal involvement. This was the basis for a shrill body of opinion following the release of the UKPDS findings that intensive glycemic control in DM 2 was not effective because the improved endpoints were not “outcomes that mattered” (eg blindness or ESRD resulting in the need for chronic dialysis). I've come to consider statements like that as anti-EBM because under first EBM principles it is the patient who decides what outcomes matter, not someone else from afar. As the JAMA paper authors acknowledge, the surrogate nature of the outcomes lasts about 8 years. Longer term follow up of the UKPDS suggested differences in more robust clinical endpoints. In patient-centered decision making a lot depends on the patient's expected longevity.

The paper goes on:

Data from randomized clinical trials consistently suggest that intensive glycemic control immediately increases the risk of severe hypoglycemia 1.5- to 3-fold. Based on these data and observational studies, for the majority of adults older than 65 years, the harms associated with a hemoglobin A1c (HbA1c) target lower than 7.5% or higher than 9% are likely to outweigh the benefits.

So according to these authors the optimal range for many patients in an internist's practice is an A1c level between 7.5 and 9. That bold statement runs counter to a lot of prevailing diabetes dogma. But the harm associated with intensive glycemic control is more than hypoglycemia. For although intensive control does without question confer microvascular benefit, it also seems to result (with a few particular exceptions) in macrovascular harm. See here.

More from the article:

However, the optimal target depends on patient factors, medications used to reach the target, life expectancy, and patient preferences about treatment. If only medications with low treatment burden and hypoglycemia risk (such as metformin) are required, a lower HbA1c target may be appropriate.

Again, it goes beyond hypoglycemia. Metformin is one of only two diabetes drugs found to confer macrovascular benefit and it is likely a pleiotropic effect, having little if anything to do with blood sugar. It would also be reasonable to say that nonpharmacologic modalities (diet, exercise) would confer benefits across the range of A1c.

The article concludes:

High-quality evidence about glycemic treatment in older adults is lacking. Optimal decisions need to be made collaboratively with patients, incorporating the likelihood of benefits and harms and patient preferences about treatment and treatment burden. For the majority of older adults, an HbA1c target between 7.5% and 9% will maximize benefits and minimize harms.

In discussions of diabetes we have, ever since DCCT and the advent of home glucometers and A1c, developed an obsession with glucose lowering. Diabetes, however, is multifaceted and there is much more to consider. Despite a few omissions, all in all this paper is a great discussion of the treatment of type 2 diabetes in accordance with the principles of evidence based medicine.

Has EBM been hijacked? Yes but not in the way John Ioannidis says it has.

Dr. David Gorski is a surgical oncologist who writes for Science Based Medicine.  I am an enthusiastic follower of his posts there and on his other blog.  I also have a passionate interest in evidence based medicine (EBM).  So when Dr. Gorski posted recently about the hijacking of EBM my interest was piqued.  Though after a careful read of this long post and some of its many links I feel I mostly agree with him, there are some points there that concern me.  Before addressing them I should provide some background.  John Ioannidis, whom Gorski cites heavily and was the inspiration for the post, is famous for his 2005 paper titled Why Most Published Research Findings Are False.  More recently he wrote the article Evidence-based medicine has been hijacked: a report to David Sackett.  In that paper, which is in the form of an open letter to one of the founders of EBM David Scakett (now deceased),  Ioannidis opines that there are widespread problems with  the medical research agenda consisting of corruption of clinical trials and investigators asking the wrong questions due to various conflicting interests.

Dr. Gorski in response is mainly favorable to the article but is concerned about a missing piece:

In his “report” to David Sackett, Ioannidis does touch on a number of pertinent and interesting points regarding the adoption of EBM but, as you will see, pretty much ignores the one huge elephant in the room.

The elephant is in reference to a popular distortion of EBM that has the effect of, well, I know of no better way to put it, enabling quackery.  The distortion in question is a tendency to devalue basic science when considering various forms of evidence.  There's a long story as to how it came about but suffice it to say here it was not the original intent of EBM's founders.  While that is the main point of Dr. Gorski's post there is much more, including some points that concern me, (and another elephant he made only indirect reference to) which I will address below.

As for Ioannidis's statement that most published research findings are false, to me it just added shock value to something we knew for decades but discussed in less loaded terms.  While he did unpack some of the reasons in a way that had not been done before we always knew that research findings are tentative and that modification of prior research by new research is the usual case.    This is something that has been acknowledged and accepted in medicine for a long time.  (I considered aspects of this phenomenon at some length in a post on medical reversal).  This is not to say we shouldn't be concerned about the quality of research.

At the risk of sounding like a stickler for correctness of terminology, while Ioannidis describes what might be called a hijacking of the research agenda it is not a hijacking of EBM.  EBM, at least as it was originally defined, is focused on how the individual clinician uses expertise to integrate the best available evidence with the needs of the individual patient at the point of care, not the research agenda.  The design and implementation of clinical trials is something separate from EBM.  Sackett himself in effect acknowledged this when in 2000 he retired from the field of EBM and migrated to the other field of clinical trials:

Dr Sackett eventually returned to Canada and, leaving the EBM field for others, devoted himself to researching and writing about randomized clinical trials in a wooden cabin on Irish Lake in Ontario. There he canoed and snowshoed with family and friends.

So while while Ioannidis points to some very important concerns what he describes is not the hijacking of EBM.  However, as Gorski points out EBM has indeed been hijacked.  I would take the discussion a step beyond what Gorski said which brings me to the other elephant in the room.  Take a look at this from his post, in reference to the discussion by Ioannidis about the profession's and industry's resistance to AHRQ:

So, yes, there is resistance to the AHRQ. However, these days it is far more business interests, such as drug and device manufacturers, than physicians groups who want to abolish both the AHRQ and the PCORI, mainly because the AHRQ and PCORI’s research threatens these companies’ bottom lines by showing which treatments work better in the “real world” and influencing the Centers for Medicare & Medicaid Services (CMS) regarding which new drugs and devices will be paid for. In fact, I’d argue that, while Ioannidis is correct that drug and device manufacturers want to kill AHRQ and PCORI, he’s missed a sea change in attitude among physicians towards such government agencies whose purpose is to evaluate and compare treatments for effectiveness after they’ve been approved. It might have been true that EBM was not popular 15 or 20 years ago, but as new generations of medical students have been inculcated with its principles and importance, EBM has been “baked in” to physician education, with a resultant change in attitude towards efforts to promote EBM. That’s not to say that physician groups don’t protect their turf. Just look at how radiologists, for example, react to new guidelines that increase the recommended age at which to start mammography or how primary care physicians react to legislation expanding the scope of practice of advanced practice nurses. However, extreme hostility to comparative effectiveness research and EBM-based guidelines has mostly retreated to fringe physician groups like the American Association of Physicians and Surgeons. Unfortunately, resistance to EBM as a constraint on physician autonomy is still fairly common, particularly among older physicians.

It gets tricky here because this paragraph is a little confusing, packing a lot in a small space and making what seem to me to be questionable assumptions.  The last sentence of the paragraph above has an embedded link to one of Gorski's old posts correctly pointing out that the notion of EBM as something that threatens physician autonomy is a straw man.  However, by equating attitudes towards government agencies with attitudes towards EBM he seems to imply that EBM promotes a top down approach to medicine.  (It's not clear to me whether that is his intended meaning).  EBM, in fact, seeks quite the opposite.  In a recent post on this very subject I quoted Sackett and some of the other founders, from one of their early articles in BMJ:

 Here's what some members of the EBM working group had to say in their seminal article in BMJ some years ago:

    Evidence based medicine is not 'cookbook' medicine...External clinical evidence can inform, but can never replace, individual clinical expertise…

    Clinicians who fear top down cookbooks will find the advocates of evidence based medicine joining them at the barricades.

The paper I quoted from is here.   In that paper we see the founders of EBM willing to fight for individual clinical judgment.  Who knew?

But getting back to another of Dr. Gorski's points, have physicians become more accepting of top down medicine in the last 10-15 years?  He seems to think they have but I don't know where he gets that.  This question isn't well informed by data but if anything the recent widespread physician outrage about ABIM, the ABIM foundation and its Choosing Wisely campaign would suggest otherwise.  (That is nicely chronicled at the blog of Westby G. Fisher, MD, FACC where most of the posts from the past two years are devoted to the topic). 

The agenda of top down medicine in its various forms as substitutes for EBM is the other elephant in the room:  EBM is being hijacked by the proponents of top down medicine.  Although folks in public policy circles don't use the term EBM very often it seems to be a widespread assumption in many of the policy discussions that these initiatives will make, even force, doctors to be more evidence based.  As policy wonk and futurist Bob Wachter once said in a discussion on top down initiatives that would stem from comparative effectiveness research:

We simply must find ways to drive the system to produce the highest quality, safest care at the lowest cost, and we need to drag the self-interested laggards along, kicking and screaming if need be.

The top down agenda is moving forward but fortunately at a slow creep and is no where near in place at the level Wachter wishes it to be.  Many of the new payment models under Obamacare (eg the ACO) are considered pilot projects and are not yet mandatory and may never be.  The AHRQ never grew the teeth that policy makers hoped it would have back when it was AHCPR.  Not all of the top down initiatives come from big government.  Many of the care pathways and performance measures offered as someone's version of EBM are locally driven.  I can only hope that doctors and medical educators will read and re-read that seminal 1996 BMJ paper and help cultivate and spread a true understanding of what EBM really is.

Should evidence based medicine be declared obsolete?

Not in my opinion but there are some policy makers out there who think it should though they won't admit it. Some by implication even falsely invoke the idea of EBM to support their agenda. These are folks who favor top down control of medicine in order to diminish the decision making power of individual doctors and patients, as Retired Doc pointed out here.

They believe that variation is the enemy of health care (remember the Dartmouth Atlas?) and that such variation is driven by the autonomy of clinicians. An example of such thinking is this quote (via the Retired Doc post) from the book “New Rules” by Drs. Donald Berwick and Troyen Brennan:

"Today, this isolated relationship[ they are speaking of the physician patient relationship] is no longer tenable or possible… Traditional medical ethics, based on the doctor-patient dyad must be reformulated to fit the new mold of the delivery of health care...Regulation must evolve. Regulating for improved medical care involves designing appropriate rules with authority...Health care is being rationalized through critical pathways and guidelines. The primary function of regulation in health care, especially as it affects the quality of medical care, is to constrain decentralized individualized decision making."

EBM, some may be surprised to learn, is opposed to this type of approach.

Here's what some members of the EBM working group had to say in their seminal article in BMJ some years ago:

Evidence based medicine is not 'cookbook' medicine...External clinical evidence can inform, but can never replace, individual clinical expertise…

Clinicians who fear top down cookbooks will find the advocates of evidence based medicine joining them at the barricades.

Clearly this is in opposition to what the policy leaders are saying. It would lend clarity to the debate if they would just be honest and say they are opposed to EBM.

The quest for open access point of care resources: can we have a free Up to Date?

Should physicians have to pay for information resources? Many physicians, among the most vocal being the pharmascolds, have maintained that they should. Recently, though, there has been a large and growing push for open access to clinical look up resources at the point of care. This was the subject of an essay and a related editorial from a recent issue of PLOS Medicine.

As the open access movement has grown increasing numbers of research articles are available free of charge on line. In addition, not only these but all Medline indexed articles can be searched for free via Pubmed, with free access to the abstracts of those journals that are not open. This, then, makes it possible for everyone to do free searching and retrieval according to the original standards of evidence based medicine. But optimal use of Pubmed requires special skills and, as I have said before, is too time consuming for most clinicians. That is one of the main barriers to the practice of evidence based medicine. The founders of EBM realized early on that the solution was to develop secondary sources, which are pre-processed summaries of clinical topics that doctors could access at the point of care. The problem is that those that have been developed, at least those widely regarded to be good, are expensive. Therein, according to the articles, lies the problem: the open access movement, for all its accomplishments, has yet to deliver pre-processed content of the type doctors need at the point of care.

It's not that it hasn't been tried. The essay cites some attempts that have failed. I can think of others not mentioned, or only given passing mention. Scientific American Medicine, for example, originated as the first continually updating paper textbook of medicine, was offered on line for free to physicians who registered with WebMD. Free access went away after a couple of years. Merck Medicus once offered free access portals to Access Medicine, MD Consult and more but they didn't last.

We still have eMedicine though it has gone through a few iterations. It remains free access and, though I have not found it as helpful as Up to Date in the past, it has improved over time.

Conspicuously absent from both papers was any mention of the open access initiatives in social media such as FOAM and BLITTER.

Conclusions and reflections:

The availability of free on line secondary source point of care information is limited.

High quality POC reference sources are largely subscription based.

Efforts on the part of the open community to build free resources have been of limited success.

Free high quality resources have mainly been available through industry support but they have not been sustainable.

There appears to be a sharp divide between pharmascolds and the open access community over whether clinicians should have to pay for access to information. Over the past decade and a half the pharmascolds have largely won the battle against industry support.

Efforts by the social media community show promise but are early in development.

The new CHEST guidelines for antithrombotic therapy

This go round the update is being handled a bit differently it would seem, by releasing one section at a time.  As of now only the section on VTE treatment has been posted, and the full text can be accessed here.  The rest of the guidelines, e.g. those for antiplatelet therapy and VTE prevention, will, I'm guessing, be posted later, a section at a time.

Retired Doc has already written a nice summary post which you can access here.

For my own take, I'll highlight the major changes and comment on some other points of particular interest.  In this edition as in previous ones the authors are true to the first principle of evidence based medicine (EBM) which is that the ultimate decision maker for treatment is the patient, providing he or she wants to be, as opposed to some pathway, core measure or central controlling authority.  This is illustrated by phrases throughout the paper, often used after their recommendation statements, to the effect that the patient may choose this over that treatment.  The document is well referenced and the recommendations seem based on careful consideration of the evidence.  Those good things being said, the paper is also a masterpiece of obfuscation, examples of which I will point out along the way.  Although there is some sense to be made of it (as I try to provide in plain language below) the process of doing so was very difficult.  The guideline is worth reading in the original text but do so when you are rested, fed and have plenty of time.


The 2012 definitions for duration of therapy are maintained.

Somewhere along the line, in 2012 I believe, the guideline authors changed the category definitions for VTE treatment duration.  They have now been around for a while but, I believe, still confuse clinicians.  The old classification seemed pretty straightforward:  short term (3 months, mainly for patients with transient risk factors), usual (6 months, appropriate for most patients with spontaneous events) and indefinite (potentially for the rest of the patient's life, for those with especially high risk of recurrence).   But the current classification turns that all around.  “Long term” is now defined as 3 months of treatment.  The middle category, “longer, time limited” is defined as 6-12 months.  The third category, termed “extended,” could be any time of longer than 3 months but without a scheduled stop date.  At the risk of pushing this to absurdity extended treatment could be as little as 3 months and one day, meaning that the order of this listing is not necessarily in order of treatment duration though it is implicit in the discussions in the paper that “extended” usually means longer than a year.


So what are the recommended durations of treatment for various clinical circumstances?

For VTE (for purposes of this post that means DVT and/or PE) not associated with active cancer and associated with a transient risk factor, surgical or otherwise, the guideline calls for 3 months treatment.  The strength of the statement (we recommend versus we suggest) varies with bleeding risk but the guideline always prefers 3 months over longer time limited or extended. 

For VTE not associated with active cancer and not associated with a transient risk factor, the order of preference for duration is extended preferred over 3 months which in turn is preferred over longer time limited (see the one exception below).  It is really difficult to make sense of this when, in plain language, it says “potentially for the rest of your life” is choice number 1, 3 months is choice number 2 and 6-12 months is choice number 3.  That is, the descending order of preference does not equal the descending order of duration!  The exception is that if the bleeding risk is high (I'll explain how the guideline defines that later) 3 months of treatment is the preference over the other two options.

To add to the confusion (and perhaps the entertainment value) of the guideline check out this statement from section 9 regarding the decision making process for patients who are potential candidates for extended (as opposed to 3 month) treatment:

Patient sex and D-dimer level measured a month after stopping anticoagulant therapy may influence the decision to stop or extend anticoagulant therapy..

But a month after stopping, well, you've already stopped!  It's almost comical when you think about it.  The plain language version would be stop treatment, wait a while, then run a test to see if you should stop.  What they really must have meant was stop treatment, wait a while, then run a test to see if you should have stopped.  It seems the authors didn't say what they really meant, perhaps because that would have sounded as though they were recommending you put the patient at undue risk.  But I do think there is a basis for this.  The body of the paper cites evidence.  I will appeal to pathophysiology and attempt to explain it this way.  While the patient is systemically anticoagulated the coagulation system is held at bay.  Once you stop, there are certain higher risk patients who will reactivate their coagulation systems in a subclinical manner, with the process simmering beneath the surface.  Such patients, identified by an elevated D dimer, would be considered poised to clot again soon.  So, in certain circumstances, the D dimer may be a useful decision tool.  But what is gender specific about it, as implied in the indented text above?  More from the guideline on that:

..patient sex and D-dimer level measured about 1 month after stopping anticoagulant therapy can help to further stratify the risk of recurrent VTE. 66-69  Men have about a 75% higher (1.75-fold) risk of recurrence compared with women, whereas patients with a positive D-dimer result have about double the risk of recurrence compared with those with a negative D-dimer, and the predictive value of these two factors appears to be additive. The risk of recurrence in women with a negative posttreatment D dimer appears to be similar to the risk that we have estimated for patients with a proximal DVT or PE that was provoked by a minor transient risk factor (approximately 15% recurrence at 5 years); consequently, the argument for extended anticoagulation in these women is not strong, suggesting that D-dimer testing will often influence a woman’s decision. The risk of recurrence in men with a negative D-dimer is not much less than the overall risk of recurrence that we have estimated for patients with an unprovoked proximal DVT or PE (approximately 25% compared with approximately 30% recurrence at 5 years); consequently, the argument for extended anticoagulation in these men is still substantial, suggesting that D-dimer testing will often not influence a male’s decision.

For patients without active cancer who have had two or more spontaneous events the recommendations are the same as for patients with a single spontaneous (ie no transient risk factor) event.

For patients with active cancer extended therapy is favored for all, although the strength of the recommendation varies with bleeding risk. 

No special consideration is given for other high risk situations (eg thrombophilia, clots in unusual places).

Finally, the guideline says that if systemic anticoagulation is stopped for any reason aspirin, absent a contraindication, is considered better than nothing, making the special point that aspirin is not an acceptable alternative for extended systemic anticoagulation.


Choice of anticoagulant:  it's complicated.

The recommendation summary is simple enough but there are many nuances.  In general, for VTE in non-cancer patients the order of preference of agents is a NOAC over warfarin.  The authors give no order of preference among the NOACs.  Concerning initiation of treatment:

Initial parenteral anticoagulation is given before dabigatran and edoxaban, is not given before rivaroxaban and apixaban, and is overlapped with VKA therapy.

There are exceptions to be noted, mainly in unstable patients, for the recommendation not to start with parenteral anticoagulants before administration of some of the NOACs listed above.

For VTE patients with active cancer LMWH is recommend over all the others for initial and long term treatment and, by implication, extended treatment.  There is no order of preference for any of the ones after LMWH (orals). 

Now for the nuances that apply to certain special situations, from table 6:

If the patient has liver disease sufficiently severe to alter the coag tests LMWH is indicated for all phases of treatment.  Two reasons are stated for this, one being that NOACs are contraindicated if the INR is raised by liver disease, and the other being that warfarin is difficult to control in such circumstances and its monitoring is confounded since when the INR is altered by liver disease it is not a reliable indicator of antithrombotic effect.

Renal impairment with clearance below 30 is considered a contraindication to all NOACs for treatment of VTE in the guideline even though the product labeling of some of them allows usage with such reduction in renal function.

Dabigatran is not recommend if the patient has CAD, dyspepsia or prior history or GI bleeding (even if remote).

Similarly rivaroxaban and edoxaban are not recommend if there has been GI bleeding.

If there is concern for poor compliance, warfarin is favored over the NOACs, since it can be detected in the lab and its consequences are not immediate.

Unfractionated heparin is favored for initial treatment if thrombolytics are used or are being considered.

I checked Up to Date regarding the choice of anticoagulant and found that it is more conservative, favoring the traditional approach over starting with a NOAC.


Distal leg DVT.

The recommendations for distal leg DVT are essentially the same as those for proximal DVT except for the acknowledgment that clinical judgment will dictate withholding treatment in some patients.  The guideline “suggests” such a strategy of withholding treatment if there are not severe symptoms or risk factors for recurrence.  In such patients serial imaging is recommended. 



Catheter directed (regional) thrombolysis is not recommended.

The authors imply that it is worthy of consideration in certain circumstances but come short of making even a soft recommendation:

Patients who are most likely to benefit from CDT (see text), who attach a high value to prevention of PTS, and a lower value to the initial complexity, cost, and risk of bleeding with CDT, are likely to choose CDT over anticoagulation alone.

An exception to this non-recommendation is “impending venous gangrene.”  Another exceptional situation, May Thurner syndrome, is not addressed.

(Note: this applies to both upper and lower extremity DVT).


The guideline finds no indication for IVC filter insertion in any patient who can be anticoagulated.

---which is another way of saying that they would recommend an IVC filter only for patients with acute VTE and an absolute contraindication to anticoagulation.  Acknowledging low level data in favor of combined anticoagulation and IVC filter insertion in patients whose PE is severe or unstable (variously defined) the guideline makes this statement:

However, because it is uncertain if there is benefit to placement of an IVC filter in anticoagulated patients with severe PE (eg, with hypotension), and this is done by some experts, our recommendation against insertion of an IVC filter in patients with acute PE who are anticoagulated may not apply to this select subgroup of patients.

Thus the guideline gives little support for IVC filter insertion in any but the strictest indications, which may be important in light of the current FDA Safety Communication on filters and the resulting flurry of legal actions.


What about compression stockings?

Although the guideline authors acknowledge that they may be useful to control leg swelling they are no longer recommended for the singular purpose of prevention of post thrombotic syndrome.


Some small pulmonary artery filling defects may not need to be treated at all.

How is this determined?  Here's the recommendation:

In patients with subsegmental PE (no involvement of more proximal pulmonary arteries) and no proximal DVT in the legs who have a (i) low risk for recurrent VTE (see text), we suggest clinical surveillance over anticoagulation (Grade 2C), and (ii) high risk for recurrent VTE (see text), we suggest anticoagulation over clinical surveillance..

As simple as this sounds it requires some elaboration.  Unanswered questions include: 1)  are they talking about a single subsegmental defect or can treatment be withheld in some cases of more than one?  2)  Is the recommendation based on the idea of a false positive (non VTE filling defect), or that such low clot burdens, even if real, don't require treatment, or both?  The discussion in the body of the paper talks around these issues but is not definitive on either one.  Suffice it to say the recommendations leave much to clinical judgment and patient preference.  Despite such ambiguity in the text here is my take:


The text contradicts itself on whether treatment can ever be withheld in real subsegmental PEs however tiny the clot burden.  However, from the explanatory text it can be inferred that all filling defects judged to be “real” should be treated since such a determination is a risk factor for progression.

Subject to variations based on bleeding risk, patient preference and other clinical judgment factors, for treatment to be withheld ALL the following conditions must be met:

The defect is single.

The CT angio leaves doubt about whether the filling defects really represent PE (ie false positive is likely).

The filling defect is distally situated in a subsegmental branch.

The patient is asymptomatic (usually meaning that the discovery of the filling defect was incidental).

Clinical pretest probability is low or intermediate.

The D dimer is normal or minimally elevated and the elevation is otherwise explained.

The patient is not hospitalized already.

The patient has no recent history of immobility.

The patient does not have active cancer.

The patient has good cardiopulmonary reserve.

DVT is ruled out by lower extremity ultrasound (the upper extremity should also be evaluated if there are risk factors such as lines).

Again this is just my attempt to make sense of the text of this portion of the guideline, which is in places ambiguous and self contradictory.  However one interprets this portion it is suggested between the lines that the authors are concerned about an emerging problem of over treatment of small filling defects, a problem that would nearly cease to exist if clinicians relied on the best evidence and chose VQ over CT scanning as the initial imaging modality to assess for PE.

A prospective observational cohort study is underway to help answer the question.


Some patients with acute PE can be treated at home.

---meaning that they can be sent straight home from the ER or sent home after only very brief (a day or so) hospitalization.  The last edition of the guidelines said that some PE patients can be sent home after a very short stay.  The current ones take it a step further to say some patients can have their entire treatment outside the hospital.  In order to be eligible the following conditions are specified in the text, all of which must be met:

1) clinically stable with good cardiopulmonary reserve; (2) no contraindications such as recent bleeding, severe renal or liver disease, or severe thrombocytopenia (ie,less than 70,000/mm 3);  (3) expected to be compliant with treatment; and (4) the patient feels well enough to be treated at home.

Note that risk scores (eg PESI), biomarkers and other clinical indicators for RV dysfunction are not included explicitly in these criteria although criterion #1 would imply that there has been some assessment of RV function and in the text the authors say that elevated biomarkers or other indicators of RV dysfunction should discourage outpatient treatment.


As in the previous edition of the guidelines, it is suggested that hypotensive PE be treated with systemic thrombolytic therapy.

---but not generally normotensive PE even when the latter is associated with RV dysfunction (aka submassive PE).  The guideline suggests that if patients with submassive PE not initially treated with thrombolysis subsequently deteriorate, even short of developing hypotension, that lysis may then be indicated.


Catheter delivered regional thrombolysis for hypotensive PE is only suggested in certain circumstances.

These are stated in the guideline thusly:

..acute PE associated with hypotension and who have (i) a high bleeding risk, (ii) failed systemic thrombolysis, or (iii) shock that is likely to cause death before systemic thrombolysis can take effect (eg, within hours)..

What about anticoagulant failure?

From the guideline text:

In patients who have recurrent VTE on VKA therapy (in the therapeutic range) or on dabigatran, rivaroxaban, apixaban, or edoxaban (and are believed to be compliant), we suggest switching to treatment with LMWH at least temporarily (Grade 2C)…

In patients who have recurrent VTE on long-term LMWH (and are believed to be compliant), we suggest increasing the dose of LMWH by about one-quarter to one-third (Grade 2C).

How is bleeding risk assessed?

Many of the recommendations and suggestions above vary with bleeding risk.  Low risk is defined as no risk factors, moderate risk is defined as 1, and 2 or more mean high.  The factors are age over 65, prior bleeding, cancer, renal failure, liver failure, thrombocytopenia, prior stroke, DM, anemia, antiplatelet therapy, poor anticoagulant control, comorbidities causing reduced functional capacity, recent surgery, frequent falls, alcohol abuse and use of nsaids.