Tuesday, June 30, 2009

Troponin measurement in the ER in patients with PE

The positive and negative predictive values for central pulmonary artery obstruction were good in this study:

Troponin values were elevated in 20 (19.2%) of 104 patients (95% confidence interval [CI], 11.6-26.8) with a mean cTnI concentration of 0.38 ± 0.44 μg/L. Elevated cTnI value had a significant correlation with main pulmonary arteries involvement using the modified Computed Tomography Obstruction Index score (P = .0001). Elevated ED cTnI value had 53.8% (95% CI, 37.6-66) sensitivity and 92.3% (95% CI, 87-96.4) specificity, 70% (95% CI, 49-86) PPV, and 85.7% (95% CI, 80.7-90) NPV for predicting main pulmonary artery obstruction on CT. Increased cTnI values were highly correlated to intensive care unit admission of patients with PE (RR, 12.83; 95% CI, 3.87-42.4).

The evidence in favor of cardiac biomarkers in the ER evaluation of suspected PE is mounting.

Monday, June 29, 2009

Salmonella infection and mycotic aneurysm

Don’t forget the association. It can be an aneurysm or a pseudoaneurysm, can occur as a new aneurysm or infection of a pre-existing one, and is associated with certain predisposing conditions. Serotype associations, according to the brief review, tend to be enteritidis and typhimurium. The article references several other reviews.

EMRs degrade the quality of clinical documentation

This article in the American Journal of Medicine explores some of the reasons. Repeated copying and pasting of other notes and template generated electronic clutter are two. Another underappreciated aspect is loss of the power of clinical narrative:

Another more insidious consequence of the copy-and-paste function has been the loss of the narrative. Because charts have become capacious warehouses of disorganized, irrelevant, or erroneous data, the story of the patient and the patient's illness is no longer easy to read or likely to be read. In a most compelling and perhaps unintended way, we are witnessing the “death” of the health record narrative, as many of us have known it. Others also speak of the loss of narrative in electronic health records, and with great concern because narratives form the basis of clinical decision making.

Daily documentation of the patient's trajectory, in prose, even when stripped of overt emotional content, is not just educational. It is humanizing.

I have yet to encounter an electronically generated note that effectively tells a patient’s story.

Saturday, June 27, 2009

Fact checking for Atul Gawande

As much as I liked Gawande's New Yorker article I'm having increasing reservations about the accuracy of his assertions, as I suggested here. I think the comment thread from Thursday's post deserves reposting:

Clinton said...
Just some fact-checking.
St. Louis County's Medicare $/beneficiary = 8,306.

Not sure if Wikipedia is a great reference, especially lacking an independent citation, but St. Louis does not show up as #3 (that spot belongs to Starr County, TX.)

Hidalgo County hits the list at #22, while St. Louis doesn't even hit the lowest 100 list. Something isn't quite right with Gawande's statistics. Maybe he is going off of a different set of measures than lowest income per capita or median household income?


R. W. Donnell said...
The figure you cite is identical to his. That's for St. Louis County, as he said. The problem is, St. Louis County is not among the poorest regions in the nation by any metric or any stretch of the imagination. The City of St. Louis (which is, I repeat, NOT in St. Louis County) may be. He doesn't seem to have any idea of what the cost per enrollee is for the City of St. Louis, but that's what he needs to cite if he wants to make his point about poverty and Medicare expenses.

My guess would be that the cost would be high in the City of St Louis. I think care is pretty fragmented and under served. Most of the hospitals (aside from ones affiliated with the two med schools) have moved to the burbs, so the picture there is pretty atypical.

I don't consider Wikipedia a very authoritative source but I know St. Louis is a city without a county from personal familiarity with the area.

Raises even more questions about his fact checking.

Oh, those greedy cardiologists

I'm all over health care variation and non-evidence based medicine (N-EBM) these days. They're hot topics because of their close ties with the current health care reform debate.

If you want some entertainment on the subject, and can stomach a little demagoguery and name calling about greedy cardiologists and pigs running the AMA by all means check out this post by Doug Bremner, M.D. If you want something factual look elsewhere.

Bremner's post is so over the top and patently absurd one wonders whether it even merits a serious response, but, evidently, some people take Bremner's blog seriously. Besides, all I have to do to smack it down is cite some simple facts, so here goes.

He starts with this:

I just found a way to save 25 billion dollars a year for President Barack Obama’s healthcare plans. That is to cut out angioplasty, for which multiple studies, including one in the June 11 edition of the New England Journal of Medicine. The mounting evidence that angioplasty is not more effective than medication treatment alone in preventing heart attack and death in people with heart disease doesn’t stop doctors from performing them.

Wrong, wrong, wrong, Dr. Bremner.

First, I'll give him the benefit of the doubt and assume he's not really talking about angioplasty but rather coronary stenting. Angioplasty as the principal coronary intervention is seldom performed anymore. In the NEJM study he cited almost all the PCI patients underwent some form of stenting.

Concerning stenting, most are not done in patients addressed in that NEJM study or the other landmark trial with similar findings, the COURAGE trial. In fact, patients with stable angina represent less than a third of those who get stents nowadays.

Let's pick apart Bremner's statement a bit more---

The mounting evidence that angioplasty is not more effective than medication treatment alone in preventing heart attack----

Huh? Cardiologists have known for over a decade that revascularization doesn't prevent heart attacks. No one is promoting it for that indication. Where does Dr. Bremner get his “information?”

Let's parse it a bit more (my italics):

The mounting evidence that angioplasty is not more effective than medication treatment alone in preventing heart attack and death in people with heart disease doesn’t stop doctors from performing them.

Nonsense. The COURAGE trial certainly did stop doctors from performing them in patients with stable angina, almost immediately:

ResultsThere was a significant increase in anti-ischemia medication use prior to catheterization referral following the COURAGE trial (mean = 1.31 [SD 0.83] medications pre-COURAGE, mean = 1.54 [SD 0.84] medications post-COURAGE, P = 0.012). Among 217 patients with coronary disease on catheterization, treatment with medication rather than percutaneous or surgical revascularization increased after COURAGE (11.1% pre-COURAGE vs 23.0% post-COURAGE, P = 0.03). There was also a significant decrease in referral volume following the COURAGE trial (3.12 referrals/day pre-COURAGE vs 2.51 referrals/day post-COURAGE, P = 0.034).

ConclusionsThe COURAGE trial immediately impacted the management of stable angina. Catheterization referral volume decreased, medication use increased, and the use of medical therapy rather than revascularization increased among patients with coronary disease.

And, there was this from Heartwire:

Use of coronary stents, including drug-eluting stents (DES), "dropped sharply" in April, the Wall Street Journal reports, citing a marketplace report conducted by Millennium Research Group in 140 US hospitals [1].

According to Journal reporter Keith J Winstein, doctors did roughly 71 200 stenting procedures in April: 10% less than in March and 15% less than the previous year. Physicians believe that drop, writes Winstein, is "an unusually quick response" to the COURAGE trial, presented at the ACC 2007 meeting in March. In COURAGE, stents (primarily bare-metal stents) were no better than optimal medical therapy at preventing future death or MI in people with stable coronary artery disease.

So here's the bomb in Bremner's post:

But I’ll give the reason why they still perform 1.2 million of these procedures every year. It is pretty simple really. Greed.

The moral preening and finger pointing that goes on in our profession is astounding.

Evidence, please, Dr. Bremner.

Perhaps the most concerning problem with Dr. Bremner's post is that he conflates stenting for stable angina with it's real evidence based use, which is in patinets with acute coronary syndromes. Emergent PCI for patients experiencing acute STEMI has been shown over and over again to save lives, save ejection fractions and get people back to work. Let's hope some misguided folks in Washington don't deprive them of it. I personally believe the doctors taking care of patients who know what they're doing should be the ones responsible for critical appraisal of best evidence, not some policy wonks from afar. See why?

Friday, June 26, 2009

Antithrombotic agents and the risk of cerebral microbleeds

Antiplatelet agents, but not anticoagulants, were associated with cerebral microbleeds in this study. While that may seem surprising at first glance it actually makes sense. Intracranial microbleed is a disease of the elderly, and is related to amyloid angiopathy in the case of lobar bleeds and hypertensive or atherosclerotic small vessel disease in the case of subtentorial bleeds. Following a tiny break in a blood vessel your first defense against such hemorrhage is the platelet plug, not the coagulation proteins. Although warfarin is known to be associated with spontaneous intracranial hemorrhage this study challenges our thinking about antithrombotic therapy in the elderly.

Amyloid angiopathy and associated hemorrhage are related in a complex way to the APO E genotype.

Medscape CME here.

Thursday, June 25, 2009

Atul Gawande answers objections

H/T to DB for pointing me to this follow up article by Atul Gawande concerning his earlier New Yorker piece on health care costs.

Here he elaborates and provides more data in response to objections and questions concerning his original article. He also re-emphasizes that it’s all about organization and leadership and gives another example, Scott and White Hospital in Temple, Texas which is part of an integrated medical group in many ways like Mayo Clinic. Despite having, purportedly, more physicians per capita than any other community in the U.S., Temple Texas has high quality scores and low costs.

One quibble. In answering the point about McAllen’s poverty as a possible driver of utilization he says:

By any measure, McAllen’s poverty and poor health fails to account for its differences from El Paso. St. Louis is located in another county that is just as poor as McAllen (it is the third-poorest county in the U.S.). Its cost per Medicare enrollee? $8,306.

St. Louis is not located in a county. It is an independent city. The surrounding St. Louis County, particularly its western aspect, is very wealthy. This leaves me wondering how well he checked his other “facts.” It is not clear whether his figure of $8,306 per enrollee represents St. Louis or St. Louis County. The demographics and culture are as different as night and day. (I grew up in the area).

Wednesday, June 24, 2009

Reasons for practice variation

In view of the reaction to Atul Gawande’s recent New Yorker article I thought it would be interesting to explore some of the many reasons for practice variation. A popular perception is that it’s a lot about greed, and Gawande provided some extreme anecdotes which suggest, on first glance, that this is the case. Objective evidence, however, suggests otherwise. Here’s a run down.

Physicians with risk averse personality profiles order more tests.

Malpractice fear drives referrals to specialists.

Internists have higher utilization than FPs according to multiple studies, attributable to being more risk averse.

I have found no study looking at greed as a driver of utilization.

QT prolongation in hospitalized patients

The horribly ill patients who are typically admitted to the hospital often have electrolyte disturbances or other conditions that may prolong the QT interval. This calls for special vigilance in drug therapy, as the list of QT prolonging drugs is daunting. This brief article from The Hospitalist has some pointers. You could just about make the case for doing an electrocardiogram on all hospitalized patients.

The electrocardiogram in pulmonary embolism

---has very poor sensitivity and only fair specificity. A new study shows that positive electrocardiographic findings indicative of right ventricular strain are additive to the prognostic information gained from echocardiography. The two techniques are complementary.

More on the over use of telemetry monitoring

This topic seems to be enjoying a resurgence of attention in the literature, perhaps as a result of its importance in bed control and the role of telemetry over use in emergency department congestion. It’s clear that telemetry use far exceeds indications deemed appropriate in the guidelines. The latest review is here.

My previous post on this topic is here.

An American Heart Association Scientific Statement which updates the 1991 ACC guidelines is linked here.

Cardiac auscultation resources from the Texas Heart Institute

There are several web based reproductions of heart sounds, most of which are of poor quality. This resource from the Texas Heart Institute is the best educational site for auscultation I’ve seen yet.

Tuesday, June 23, 2009

Neuro-imaging tips for hospitalists

Here's a summary of one of the neuro talks at Bob Wachter's Hospital Medicine course last fall. It contains a lot of pearls.

Monday, June 22, 2009

Should TPN be started and stopped gradually?

When I was in training TPN was exclusively within the purview of the surgeons. Consequently, internal medicine residents were not trained in its administration. Despite a lack of training and experience hospitalists are increasingly being asked to take over this aspect of hospital care.

One of the popular maxims regarding TPN was, because of the risk of severe dysglycemia, never to start or stop abruptly. Myth or fact? This Medscape Ask the Experts piece addresses the question.

The importance of family history in assessing the risk of first venous thrombosis

For a good while we’ve known about the importance of family history in the assessment of patients’ risk for VTE. In many ways the family history is more powerful than all the expensive thrombophilia tests. The more common genetic thrombophilias have low penetrance rates. Thus, when thrombophilia is expressed clinically in families, two or more genetic risks are often present in combination. This helps explain why the family history is often more helpful than laboratory testing.

Another factor is the state of the science of thrombophilia testing. Take 100 patients with familial thrombophilia and do an extensive battery of genetic and phenotypic laboratory tests. You’ll find something in only around 60 of them. The other 40 have a genetic disorder waiting to be discovered. Back in the 1980’s, when the relatively uncommon anti-thrombin, protein C and protein S deficiencies were the only hereditary thrombophilias known the number of patients with a positive test would have been around 15. The number has increased only recently with the discovery of the most common thrombophilias, factor V Leiden and the prothrombin mutation. (The more common thrombophilias tend to be discovered later because of their low penetrance).

A population-based case-control study recently published in the Archives of Internal Medicine adds to our understanding of the risk factors for VTE and their interactions. It demonstrated that family history was at least as predictive as, and correlated poorly with, known genetic risk factors. The full text, if you can access it, is well worth the read as it provides a world of background information which should be helpful to the clinician in understanding the optimal use of thrombophilia testing.

Sunday, June 21, 2009

The absurdity of Medicare's never events policy

A commentary in NEJM cites evidence that at best only 20% of hospital falls, a currently designated “never event,” can be prevented. Concerning VTE, under consideration for addition to the list, Aggravated DocSurg provides evidence that some hypercoagulable patients are gonna clot despite appropriate VTE prophylaxis.

In its report on patient safety a decade ago the Institute of Medicine emphasized the need to move from a culture of blame to a culture of transparency. Instead, because of the never events initiative we've derailed transparency and taken the culture of blame to a new level. The patient safety movement has backfired. We're worse off than we were 10 years ago.

H/T to Retired Doc.

Atul Gawande on health care costs

Bob Wachter, whose blog post first alerted me to Atul Gawande's brilliant New Yorker piece, notes that it is essential reading for anyone concerned with health care policy. Gawande has a unique and fascinating way of explaining health care variations. What he does in the article is explain Dartmouth Atlas data on health care variations by telling stories---stories about the organization (or disorganization) of health care delivery, mainly in McAllen, Texas, but also in places like the Mayo Clinic and Grand Junction Colorado.

The strength of Gawande's account is that it is vivid and experiential. (As you read it reflect on what's going on in your own medical community). Its major limitation is not only that it is anecdotal, but that its anecdotes represent the extremes in health care variation---only the best and worst in the U.S. are profiled. Although Gawande makes no pretense of offering a rigorous, comprehensive analysis of the problem his article has already caught the attention of policy makers, who may be tempted to overgeneralize. Concerning the piece Wachter drops this bomb:

He swats away the usual explanations (our patients are sicker, more obese, more addicted, more Mexican; our lawyers are nastier; our quality is better…) to unblinkingly zoom in on the real culprit: a culture in which providers’ greed trumps the patients’ interests.

I hope the medical education thought police don't run with that idea---next thing you know greed management classes will be required CME for doctors. (Hold on a minute while I search the web for a Dartmouth greed map.................Nope. Couldn't find one). Well, if a national greed map existed I think it would be homogeneous. It's an inherent human trait, such a constant, that what plausible explanation could there be for selective expressions of greed among health professionals in geographic regions?

Even if Bob's cynical explanation applies partially to Gawande's extreme anecdotes those “usual explanations” are not so easily swatted away when one looks at the big picture of health care in the U.S. (Juxtapose a national obesity map on one of the Dartmouth Atlas maps and you'll see what I mean).

Gawande, avoiding pat answers about greed, takes a nuanced look at complex cultural factors in the organization and delivery of health care. The Mayo Clinic figures prominently in his story. Rochester Minnesota, whose residents get almost all their care from the Mayo system, ranks in the bottom 15% for health expenditures. That's difficult to understand from a simplistic view of utilization. Mayo doesn't skimp on the best medicine has to offer, as anyone who has dealt with their system knows. Unfortunately public policy leaders who take a superficial view of greed driving health care costs may see the solution as merely one of denying services. But, as I noted before, the best research we have on practice variation indicates that underutilization is a greater problem than overutilization.

The real key, in Gawande's analysis, is organization and design. Mayo is a lesson in how efficiency and quality can be designed into a local system. Gawande observed it first hand during a visit there:

Among the things that stand out from that visit was how much time the doctors spent with patients. There was no churn—no shuttling patients in and out of rooms while the doctor bounces from one to the other. I accompanied a colleague while he saw patients. Most of the patients, like those in my clinic, required about twenty minutes. But one patient had colon cancer and a number of other complex issues, including heart disease. The physician spent an hour with her, sorting things out. He phoned a cardiologist with a question.

“I’ll be there,” the cardiologist said.

Fifteen minutes later, he was. They mulled over everything together. The cardiologist adjusted a medication, and said that no further testing was needed. He cleared the patient for surgery, and the operating room gave her a slot the next day.

The whole interaction was astonishing to me. Just having the cardiologist pop down to see the patient with the surgeon would be unimaginable at my hospital. The time required wouldn’t pay. The time required just to organize the system wouldn’t pay.

According to Wikipedia:

Mayo Clinic pays medical doctors a fixed salary that is unaffected by patient volume. This practice is thought to decrease the monetary motivation to see patients in large numbers and increase the incentive to spend more time with individuals.

That doesn't mean Mayo Clinic doctors are less greedy. They are paid handsomely. How else would Mayo recruit and retain the best and the brightest? What it does reflect is an enlightened view among the leaders about physicians' incentives.

Gawande talked with Mayo's CEO about how better alignment of incentives promotes efficiency:

No one there actually intends to do fewer expensive scans and procedures than is done elsewhere in the country. The aim is to raise quality and to help doctors and other staff members work as a team. But, almost by happenstance, the result has been lower costs.

“When doctors put their heads together in a room, when they share expertise, you get more thinking and less testing,” Cortese told me.

How can Mayo clinic be efficient and patient centered at the same time? If they had a motto for patient centeredness it might read: apply the best available knowledge and resources to offer patients what they need, no more and no less, and eliminate all conflicts of interest and institutional barriers that stand in the way. That, in contrast to Don Berwick's definition (give patients what they want, how they want it and when they want it), is a vision for patient centeredness that doesn't trump efficiency.

Gawande goes on to profile Grand Junction, Colorado, another community with low health expenditures. The medical community there is not structured like Mayo but has in common a level of organization and integration that promotes efficiency.

What's the take-away message from Gawande's article? That it's not just about greed. Greed is part of all of us, but it can't account for regional variations. It can be managed if the incentives are properly aligned. That's why organization is key. That it's not about who pays for it, nor is it about nationalization. He writes, in conclusion (my italics):

But a choice must be made. Whom do we want in charge of managing the full complexity of medical care? We can turn to insurers (whether public or private), which have proved repeatedly that they can’t do it. Or we can turn to the local medical communities, which have proved that they can. But we have to choose someone—because, in much of the country, no one is in charge. And the result is the most wasteful and the least sustainable health-care system in the world.

Organization and integration at the local level. Even if it's one community at a time. It won't be a quick fix, but we can make a difference if we start now. That's the message.

Friday, June 19, 2009

Roy Poses on EBM and CER

Roy Poses M.D. is one of the bloggers over at Health Care Renewal. He's no stranger to EBM and the related issues of helping physicians keep up with the onslaught of scientific information and helping them put best evidence into clinical practice. I have participated with him in several Medscape Roundtable Discussions. Although we sometimes disagree on areas of public policy in medicine, particularly the role of industry, I respect his views.

Recently he has written one of several posts on comparative effectiveness research (CER). Like many opinion writers on this subject he conflates CER (which is nothing more than head to head clinical trials, which we've been doing for decades) with an agenda for more government control of research and health care. That conveniently enables him and other writers to accuse skeptics of this advancing agenda of being opposed to CER, a fallacy which I exposed here. Any defense of CER, as Dr. Poses purports to be making, presupposes that somewhere there's opposition. But, as I said in that post, there is no serious opposition to the pure notion of CER.

The first example he cites is a piece by Robert Goldberg from almost 2 years ago, originally published in the Washington Times. There Goldberg ranted against the government agenda to control costs and limit health care choices and criticized the ALLHAT study. As Poses points out the article was distorted and a little over the top, but while Goldberg spoke derisively about the “comparative effectiveness crowd” and the “evidence based crowd” nowhere in the article did he say we shouldn't apply the best evidence to individual patient circumstances, and nowhere in the article did he call for a moratorium on head-to-head clinical trials conducted in real world situations. (I have criticized the EBM crowd myself although I love the pure notion of EBM). I don't know what Goldberg really feels about the pure notions of EBM and CER. He didn't say in the article.

The next example is this op-ed in the New York Times in which Peter Pitts wrote:

...the provision would allocate $300 million to create a Center for Comparative Effectiveness that would test whether newer, more expensive drugs work better than their older and cheaper counterparts. Medicare would use the center’s findings to help decide which drugs to cover. If the center found that a newer, pricier pill was no more effective than the older, cheaper version, Medicare would probably refuse to pay for it.

This sounds reasonable. But it will most likely result in Medicare covering fewer breakthrough medicines, which would, in turn, force doctors to prescribe only the drugs that Medicare will pay for — not the ones that are best for the patient.

Again, Pitts is against government control of the research and the likely agenda to surrounding it, not the research itself. In the same way Dr. Poses has been critical of some research agendas. He once wrote:

Vested interests may also try to manipulate the design, implementation, and analysis of research studies to increase the likelihood that results will be favorable to them. Some possible manipulation tactics were noted by Smith and Brophy. Such manipulation has been made easier by the surprising willingness of US medical schools and academic medical centers to let commercial sponsors, rather than their own faculty "investigators," control clinical research studies.

Like Pitts, Dr. Poses is not arguing against any particular type of research, but he is concerned, like Pitts, about who pays for it and who controls it.

Dr. Poses points out one of the shortcomings of EBM in seeking to apply the best available evidence: the best available evidence varies in quality from one clinical problem to another. For some conditions we have very good evidence, but for others it is lacking or of poor quality. We all agree on the need more and better evidence. No one is arguing for a moratorium on head-to-head clinical trials or any other type of research.

But later in his post, purportedly in support of his “defense” of CER, he quotes a New York Times piece by Tyler Cowan which deals with health care spending but is not about CER at all. Here's some of what Cowan says:

Drawing upon the ideas of the Harvard economist David Cutler, the Obama administration talks of empowering an independent board of experts to judge the comparative effectiveness of health care expenditures; the goal is to limit or withdraw Medicare support for ineffective ones. This idea is long overdue, and the critics who contend that it amounts to “rationing” or “the government telling you which medical treatments you can have” are missing the point.

Incidentally, note that Cowan doesn't dispute the idea of government wanting to ration and tell us what treatments we can have, he just thinks it's beside the point. But Cowan isn't calling for more research. He's talking about a government panel to judge the findings of existing research evidence. Because he blames the problem on doctors' financial incentives he wants to take an essential step of EBM, critical appraisal, away from the treating doctors and put it in the hands of a government panel! That's what the naysayers are riled about. Will that help the cause of EBM? Is it what Dr. Poses really wants? (Take a look at the abysmal failure of what the government panels have handed us so far).

Cowan goes on to make this interesting observation:

Of course, we have not made such Medicare spending cuts yet, and there are few signs that we will. A Kaiser Family Foundation poll found that 67 percent of Americans believe that they do not receive enough treatment and that only 16 percent believe that they have received unnecessary care.

That's uncannily close to what the best research on the appropriateness of health care utilization shows (!):

We also classified indicators according to the problem with quality that was deemed most likely to occur, and we found greater problems with underuse (46.3 percent of participants did not receive recommended care [95 percent confidence interval, 45.8 to 46.8]) than with overuse (11.3 percent of participants received care that was not recommended and was potentially harmful [95 percent confidence interval, 10.2 to 12.4]).

Put another way, if doctors were 100% adherent to evidence based practice expenditures would more likely rise than fall. Government policy makers are smart, and surely they know this, but they don't want the rest of us to know. So much for transparency. The government is extremely conflicted in their research agenda, every bit as much as the private interests. They want to convince us that cheaper is better. Given the research that shows otherwise it will take some creative design and interpretation to support their agenda.

Pearls on toxicology

---and a vast array of other topics were presented at the American College of Emergency Physicians Scientific Assembly last fall. Leon Gussow, MD, an assistant professor of emergency medicine at Rush Medical College, wrote in Emergency Medicine News:

A great many informative and entertaining lectures were presented at the American College of Emergency Physicians' Scientific Assembly, all essential to the practice of emergency medicine, but here I examine some important issues raised in lectures related to medical toxicology.

These lectures contain some fascinating content for seasoned clinicians, such as you’re unlikely to find anywhere else. You can access the pdf files of the PowerPoint presentations for the toxicology content and all the other lectures here.

Since this CME accredited meeting is awash in industry support (300 companies or so) it serves as a good example of what we’ll lose following the ban on commercial support for CME. It also provides an opportunity for the Dr. RW challenge to the PharmaScolds of the world to examine the meeting’s content, call out any biased or distorted information and convince skeptics like myself that this type of meeting should be banned.

Acute coronary syndrome management

Recent review in NEJM.

Thursday, June 18, 2009

More on the cardiotoxic effects of methadone

There has been some controversy about methadone’s cardiac risk along with vigorous objections to the ACP guidelines for cardiac safety monitoring, coming mostly from advocates for methadone maintenance treatment for opiate addiction. My last post on the subject drew a blistering reply. The commenter is the author of this blog. She wrote:

Your broad assumption about methadone's cardio toxicity has yet to be proven and your talking about it like it's a given? It never fails to surprise me how easily even DOCTORS can leave their rationality and objectiveness behind, when they hear the word "methadone".

YOur writing this entry as if your assumptions about methadone are a "known fact" and it's simply not true.

Although there is widespread ignorance concerning methadone’s cardiac toxicity no informed person on either side of the debate, that I know of, denies the toxicity. What is in dispute is the degree of risk as well as whether, and how, the evidence of this risk should change practice. The evidence spans multiple levels from molecular mechanisms to clinical observations.

It’s been a while since I’ve compiled the evidence in one place so here’s a summary:

Methadone blocks the potassium channel.

An early report of torsade associated with very high doses of methadone.

Torsade and QT prolongation were common in hospitalized patients receiving methadone.

Sudden cardiac death was attributable to methadone in patients who had therapeutic (non-toxic) blood levels.

She went on to say:

The first study you mention was very poorly done, only has 17 participants and everyone in it had a dose of 400mg or more. Hardly what most people reading your blog are going to think of when they read "ordinary therapuetic dose".

That particular study (the second one linked above), a paper in the Annals of Internal Medicine, was not the basis for my comment about methadone causing cardiac problems in ordinary therapeutic doses. That comment was linked to this study (the last one linked above) showing that patients who experienced sudden cardiac death had therapeutic levels of methadone.

And this:

The recent rise in "body count" due to methadone has to do with OVERDOSES, not sudden cardiac death--and is usually in pain patients who took too much or in addicts who are not in a methadone program and they are almost ALWAYS poly drug...so bringing up the rise in methadone RELATED deaths is a poor choice. People will read your entry and believe that we have thousands of people dying because they took a small dose of methadone and their heart exploded. This is simply not the case.

Here’s a graphic from the CDC on the rise of deaths attributable to methadone as opposed to other drug poisonings.

Data from that same CDC report suggest that overdose is not the primary cause of rising deaths:

Since 1999, between 73 and 79 percent of poisoning deaths mentioning methadone have been classified as unintentional (3,202 such deaths in 2004), with an additional 11-13 percent being of undetermined intent, 5-7 percent as suicides, less than 1 percent as homicides, and about 1 percent were injuries other than poisoning.

73%-79% were classified as unintentional, not attributable to suicide or homicide. It could be argued that those deaths involved excessive use (and the CDC does refer to them as poisonings) but absent suicide or homicide what’s an overdose? Any death attributable to a drug is, in a sense, an example of excessive use. For opiates, overdose is judged by individual circumstances, there being no generalizable quantitative definition.

The commenter went on:

Posting the comment from another site about a doctor knowing methadone is to blame immediately by looking at the EKG. First, the symptoms you mention (pinpoint pupils and lethargy) happen in any overdose situation.

She was referring to my quotation from the article by James R. Roberts, MD:

The prescient clinician faced with a lethargic patient with pinpoint pupils can glance at the EKG and immediately pronounce methadone overdose.

First of all, pinpoint pupils do not “happen in any overdose situation.” If that were true pupillary examination would be of no value in identifying specific toxidromes. As it happens, some poisonings are associated with dilated pupils, such as TCA overdose. Pinpoint pupils do not prove methadone overdose, to be sure, but the electrocardiogram can be of added value. In a lethargic or comatose patient the combination of pinpoint pupils and long QT does strongly implicate methadone. Other factors that prolong the QT can be readily excluded (check the patient’s K+, Mg++ and look at their med list). TCA overdose, for example, is suggested by dilated pupils, a widened QRS and AVR pointing to the sky! There are in fact several electrocardiographic toxidromes. The whole point is that when presented with a comatose patient a rapid low tech assessment can provide powerful diagnostic information.

There is some truth in this point:

To date there has been no proof that routine EKG's, for anyone on methadone, will even HELP--and it may actually hurt many patients because of the cost. Especially considering that there is no proof that RANDOM ekg's will diagnose or help prevent this phenomenon.

True except for the cost, which is miniscule. The guidelines call for zero risk interventions. They are nothing more than drug safety monitoring recommendations. Drug safety monitoring is supported by post marketing reports, physiologic rationale and clinical judgment. Few if any drug safety monitoring recommendations are validated by RCTs that prove the benefit of the monitoring. If such high level evidence were required most FDA safety recommendations would have to be retracted.

Several letter writers expressed concern that the guidelines would produce barriers to appropriate use of methadone for patients in need of treatment for opiate addiction. But the guidelines only recommend precautions. They do not prohibit methadone use in any circumstance. In fact, failure to observe safety recommendations for methadone could result in it being yanked by the FDA---maybe not for the critically important niche of opiate addiction treatment, but a ban of use for chronic pain could occur as more reports trickle in. (That’s exactly what happened to cisapride, a drug known to cause torsade but which might still be on the market today had docs followed the warnings).

By the way, the guideline writers at the ACP aren’t the only ones concerned about methadone’s cardiac risks. The FDA issued this warning and the Center for Education and Research on Therapeutics, the leading repository of QT prolonging drugs, has placed methadone in the highest risk category.

Problems with IE 8 and Blogger

Microsoft has recently pushed out IE 8 as an update for WinXP users. I did the “upgrade” this week from IE 6 to IE 8. The browser has that nice new look and seemed to work great---until I tried to paste something into Blogger’s post editor. No can do!

Firefox is my work around for now. Here’s a discussion of the problem in Google’s help forum.

Acid suppression and nosocomial pneumonia

A recent JAMA paper only confirms the long known association:

Using multivariable logistic regression, the adjusted OR of hospital-acquired pneumonia in the group exposed to acid-suppressive medication was 1.3 (95% CI, 1.1-1.4). The matched propensity-score analyses yielded identical results. The association was significant for proton-pump inhibitors (OR, 1.3; 95% CI, 1.1-1.4) but not for histamine2 receptor antagonists (OR, 1.2; 95% CI, 0.98-1.4).

Conclusions In this large, hospital-based pharmacoepidemiologic cohort, acid-suppressive medication use was associated with 30% increased odds of hospital-acquired pneumonia. In subset analyses, statistically significant risk was demonstrated only for proton-pump inhibitor use.

The recent emphasis on the ventilator bundle has created widespread confusion on this point. The bundle (sometimes known as the VAP bundle) has been touted as a process improvement to decrease the incidence of ventilator associated pneumonia, despite the fact that acid suppressive medication is one of the bundle components, and such treatment has long been known to increase pneumonia.

(Elsewhere I pointed out the lack of evidence in support of the bundle as a means to decrease pneumonia).

Many hospitalized patients are on PPIs for weak and sloppy indications. Better stewardship in the use of PPIs may help reduce nosocomial pneumonia.

Wednesday, June 17, 2009

Blogging has suffered lately

---due to a very punishing work load last week. Like many hospitalist programs these days ours has a 7 on 7 off schedule. I try and store some posts in the can during my week off for posting later. Were it not for Blogger's auto-post the Notes would have shut down last week. Can't blog at work and too tired to blog when I get home.

Although auto-posting continued apace I had too little time to devote much attention to some exceptional posts elsewhere in the blogosphere. Although it will seem late I intend to comment in some detail on these topics over the next few days:

Atul Gawande’s important New Yorker article as discussed by Bob Wachter and DB.

Dr. Wes's Open Letter to Patients Regarding Health Reform. (Well, I may not need to comment. This masterpiece needs no elaboration. But do go and read it).

Commentary by Retired Doc and Aggravated DocSurg on the absurdity of Medicare's “never events” policy.

Commentary from Retired Doc and other sources on Obama's speech to the AMA.

Now how many of these will I actually get to?

Can in-hospital code blues be anticipated and prevented?

We know that rapid response teams (RRTs) and telemetry monitoring have been disappointing in this regard. Telemetry monitoring is known to be overutilized. Although telemetry guidelines, published in 1991, are no longer available on line, here is a modified version of those guidelines adopted by Jackson Memorial Hospital, Miami Florida.

Why have these tools been ineffective? While telemetry monitoring can alert staff to a code in progress (thus theoretically improving the resuscitation outcome) it is of limited usefulness for anticipating cardiac arrest, which is usually heralded by deterioration of multiple clinical parameters besides cardiac rhythm.

The disappointing results of RRTs have been subject to much speculation. Physiologic deterioration often precedes cardiac arrest by 24 hours or more. Given that RRTs focus on an “emergency” response triggered at the last minute they may represent inappropriate substitutes for clinical vigilance. This fact has led to heightened interest in clinical tools such as the early warning score (EWS) and the modified early warning score (MEWS). Although research results on the EWS and MEWS have been mixed, a recent paper in the International Journal of Clinical Practice (via Medscape) was encouraging:

In comparison with the lowest score, the risk of death was incremental among all the MEWS categories, as well as the risk of the combined outcome of death and transfer, and highly significant…

We have confirmed that the MEWS, even when calculated once on admission, is a simple but highly useful tool to predict a worse in-hospital outcome.

Patients with MEWS scores of 2, 3, 4 and 5 or greater had a risk of in hospital death or transfer to higher acuity of care of 17.5%, 26.8%, 36.4% and 43.6% respectively.

Here’s a quick link for the MEWS calculation.

Tuesday, June 16, 2009

Unique aspects of methadone toxicity---the methadone toxidrome

James R. Roberts, MD, Professor of Emergency Medicine and Toxicology at the Drexel University College of Medicine, describes what I think we should call the methadone toxidrome:

The prescient clinician faced with a lethargic patient with pinpoint pupils can glance at the EKG and immediately pronounce methadone overdose.

Although the cardiac toxicity of methadone, unique among analgesics, has been known for at least 7 years and I first blogged about it here, it remains underappreciated. While Sid Wolfe clamors to ban Darvon the rising methadone body count continues to be ignored by consumer activists. All the while the pain treatment mafia rails against the evils of Demerol while promoting methadone as a reasonable analgesic.

Methadone’s cardiac issues first surfaced on a hospice which employed heavy use of methadone for terminal analgesia where otherwise stable patients, instead of gradually slipping away as expected, were dying suddenly. The same phenomenon has apparently been observed in methadone maintenance populations. To be sure, cardiac death can be the mode of demise in methadone overdose (the cardiac effects are not reversed by narcan, by the way) but not only that. Methadone can cause sudden cardiac death in ordinary therapeutic doses. And while the respiratory depressant effects of methadone, along with the analgesic effects, are blunted by the gradual development of tolerance the cardiotoxic effects are not. This is why patients on methadone may experience cardiac problems after years of stable use. Indeed, as dosages increase with the development of tolerance the cardiac risks can only increase over time.

Dr. Roberts seems to have gotten one thing wrong in this otherwise excellent article. He says:

No one appears to recommend routine EKG or continuing EKG follow-up in patients undergoing MMT or for those treated with methadone for chronic pain.

That was true until early this year when the American College of Physicians issued cardiac screening and safety monitoring guidelines for patients being prescribed methadone.

Monday, June 15, 2009

In case you need a refresher on CPR and ACLS

---here’s yet another review. This Medscape linked full text article discusses the 2005 guidelines, the evidentiary and basic science rationale behind them and some research reported since publication of the guidelines. Emerging evidence is suggesting that compression only resuscitation may be superior to compression-ventilation resuscitation for many types of cardiac arrest.

Should CPR be done before defibrillation? In an unwitnessed arrest, absolutely. For a witnessed arrest it depends on the immediacy with which defibrillation can be applied. The review discussed the rationale and related controversies.

Therapeutic hypothermia is discussed. Despite substantial evidence in favor of applied hypothermia post cardiac arrest the uptake of this evidence is low. The article notes:

Despite this evidence, studies have found that the majority of physicians have not used induced hypothermia in the management of cardiac arrest survivors. Future studies are required to investigate the barriers and solution to instituting induced hypothermia.

Here’s my not so humble opinion on what the barriers are.

Hypertonic saline for post traumatic shock not beneficial

The NHLBI trial has been halted.

We’ve been trying for decades, but we still can’t seem to find any resuscitation fluid that’s better than good old isotonic saline.

Friday, June 12, 2009

The extended time window for IV TPA in acute ischemic stroke

---is now “official” in the U.S.

Medscape report here.

Update from AHA/ASA (new Science Advisory) here.

2007 acute ischemic stroke guidelines here.

Primary source (ECASS trial) here.

NEJM editorial here.

The importance of lead AVR

The most neglected electrocardiographic lead may be the most important in some patients with ACS. From Medscape Emergency Medicine Viewpoints:

The takeaway point is simple: When patients with ACS, including non-STE ACS, demonstrate STE in lead aVR, the aggressiveness of early management must be
increased. These patients have more complex coronary lesions and will likely benefit from earlier invasive therapy.

Background here.

The Lake Wobegon effect in the treatment of hypertension

Not all docs are above average but, according to this study, most of us tend to think we do a better job of treating HT than we actually do. Moreover, when the patient doesn’t reach goal we blame it on “noncompliance.”

Is it a case of pure self deception or are there external factors at work? Lack of familiarity with the HT guidelines and the barriers inherent in treating all of patients’ multiple problems to “goal” were cited in this Medscape report on the study:

"The JNC guidelines are 60 to 70 pages long. In primary care, we treat a vast variety of diseases, all of which come with their own guidelines, and we then run into the problem of competing demands. The reality is that if primary-care physicians did all the screening and preventive recommendations that are out there, there would be about 20 to 30 minutes left in the day to do everything else. So it becomes a matter of focus. And in our attempts to focus on a lot of things, we're not seeing the forest for the trees," said Wexler.

Or, as DB recently put it, when patients have multiple problems we have to prioritize:

We are unlikely to do enough research to develop guidelines for each situation. As physicians we must make difficult decisions about which diseases to treat aggressively and which diseases deserve less aggressive measures.

On a related note one expert interviewed for the Medscape piece said that doctors would be more likely to embrace guidelines calling for different targets in different comorbid conditions if they were supported by more research on what targets are best:

In answer to Wexler's plea for more concise recommendations for the treatment of blood pressure, Nicholls said more evidence-based studies to answer the "really important question--what is the ideal blood pressure that our patients at various levels of risk should be at?" are essential.

"If we do those studies, and we get more and more evidence, the blood-pressure guidelines will become a lot clearer, and then, I think, clinicians will feel that they are
in a much more comfortable position to be able to embrace the guidelines and follow them."

What‘s better, 140/90, 130/80 or 120/70? Ask those questions for each of several comorbidities and age ranges. We’re doing that with lipid control now (where the field has reached some degree of maturity), ambulatory diabetes control (where we have more questions than answers) and in patient glycemic control (where our knowledge is in its infancy). Think of the research possibilities! It could be a whole new agenda: comparative target research!

The parallel universes of clinical documentation and administrative documentation

When it comes to honest to goodness clinical documentation doctors are pretty good. But when it comes to administrative “clinical” documentation it’s a different world, and docs need some help. Administrative documentation is based on the decades obsolete ICD 9 codes and is used to determine hospital reimbursement.

Via ACP Hospitalist blog.

Thursday, June 11, 2009

The problem with EMRs

Despite public clamor for widespread adoption since publication of the IOM’s To Err is Human, a decade later we still lack hard clinical data to show that EMRs enhance patient safety. If anything they may be adverse to the cause of patient safety. Why? Multiple reasons are considered in a recent editorial in the Archives of Internal Medicine. In particular the authors list nine pertaining to CPOE:

(1) additional work for clinicians; (2) unfavorable workflow changes; (3) never-ending demands for system changes; (4) problems relating to persistence of paper records; (5) changes in communication practices with false assumptions; (6) negative emotions generated from changing established practices; (7) generation of new types of errors ("e-iatrogenesis"12); (8) loss of ordering autonomy to accommodate CPOE goals and system limitations; and (9) overdependence on the new technology.13

For the most part these are problems in the institutional culture built around the EMR rather than any limitations of the EMR itself. Additional work for clinicians is created by taking nurses and secretaries out of the loop of many aspects of order processing in which clinicians have no prior training or experience. The theory is that removing a layer of human involvement removes one more opportunity for error, making the ordering more direct and clean. That theory, by the way, has no evidence to support it, and it’s equally plausible that this actually removes a layer of safety. In other words maybe we would be better off to take the P out of CPOE! In many ways order processing is baggage which distracts doctors from real clinical issues.

Item 5 is “changes in communication with false assumptions.” You bet. One such false assumption is that after conversion to a paperless system nurses will remain as engaged as they were in the paper days. Not that they don’t want to stay engaged (and the exceptional ones will), but the EMR burdens them with so much onerous and meaningless “documentation” that their time for honest to goodness patient care is limited.

Item 8 speaks of “loss of ordering autonomy to accommodate CPOE goals and system limitations.” Yeah. Remember how nice and easy it was to write your favorite 3 day steroid taper in paper orders? Or “hold coumadin any time INR is greater than 3?”

Electronic notes open up a whole different set of problems:

For example, rigid structures or templates impede readability, and the patient's story may become a patchwork of cut and pasted excerpts.19 Cut, pasted, and propagated
preliminary test reports or inaccurate historical content can easily lead to incorrect decisions downstream. Independent histories and examination findings may never be recorded or may be missed as when a single physician's history is copied again and again. Overly long notes, facilitated by "copying forward" with new daily increments, contribute to an inability to quickly page through the medical chart for critical information…. It becomes clear quickly why a loss of confidence in the accuracy of the medical chart might emerge.

I’ll take the old handwritten barely legible one liner any day.

The mandate for universal adoption of EMRs over the next several years is ill conceived and premature.

The VIPoma syndrome

Reviewed in Hospital Physician.

When is thrombolytic therapy indicated in PE

This review says pretty much the same thing all the others say. Assuming no contraindications:

Hemodynamically stable, no RV dysfunction---no.

Hemodynamically stable, RV dysfunction---maybe.


Cardiac arrest suspected due to PE---what have you got to lose?

Which ischemic stroke patients need a TEE?

The main reason for getting a TEE in a patient with ischemic stroke is to find out if that patient has an indication for warfarin anticoagulation. You can begin the selection process by excluding patients who already have an obvious indication for warfarin (e.g. atrial fibrillation) and those who have a contraindication for warfarin.

Among the remaining patients only those with cryptogenic stroke are likely to need a TEE. I previously linked to a paper demonstrating a high yield for TEE in finding an indication for warfarin in such patients. Here is a new study documenting similar findings:

Results: 702 consecutive patients (380 male, 383 IS, 319 TIA, age 18–90 years) were included. In 52.6% of all patients, TEE examination revealed relevant findings. Overall, the most common findings in all patients were: patent foramen ovale (21.7%), previously undiagnosed valvular disease (15.8%), aortic plaques, aortic valve sclerosis, atrial septal aneurysms, regional myocardial dyskinesia, dilated left atrium and atrial septal defects. Older patients (greater than 55 years, n = 291) and patients with IS had more relevant echocardiographic findings than younger patients or patients with TIA, respectively (p = 0.002, p = 0.003). The prevalence rates of PFO or ASD were higher in younger patients (PFO: 26.8% vs. 18.0%, p = 0.005, ASD: 9.6% vs. 4.9%, p = 0.014).

All in all, based on these two studies, a third to half of patients with cryptogenic stroke are found to have potential indications for warfarin anticoagulation.

Background: The TOAST classification may be helpful in the determination of cryptogenic stroke.

Wednesday, June 10, 2009

RAC attack!!

Medicare, to the extent that it doesn’t pay for the care you are mandated to deliver, is a provider tax. Of late the folks at Medicare have decided that your tax refund in the form of Medicare fees may be excessive. So Medicare has contracted with “bounty hunters” in the form of Recovery Audit Contractors (RACs). This article from the Journal of American Physicians and Surgeons explains the program. Here are a few tidbits:

Medicare’s recovery audit contractor (RAC) program, the “bounty hunter program,” is now a permanent entity.

As the Medicare program is facing $34 trillion in unfunded liabilities, it is clear that government has promised more in Medicare benefits than taxpayers can afford long-term.

Those physicians who are holding out for a “fix” in the flawed Medicare SGR
(sustainable growth rate) payment formula should know that if a “fix” is implemented, it will likely come at the cost of the adoption of a DRG-like system of payment for outpatient encounters (episodes of care). The adoption of a DRG-like system of payment for outpatient encounters will, of course, ensure that patients who present to the physician’s office will be given the same type of treatment that patients receive in the hospital under the DRG (diagnosis related group) payment system.

(Just such a DRG-like system for the out patient sector was proposed in 1985 in the form of the Kennedy-Gephardt bill, which was defeated).

Here’s more:

Medicare is a giant Ponzi scheme that, like all such schemes, is destined for collapse. Despite repeated warnings of impending financial collapse by AAPS, the former head of the Government Accountability Office (GAO), the former Secretary of the Department of Health and Human Services, and the chairman of the Federal Reserve, the giant Medicare Ponzi scheme continues, and soon will take on the additional costs of the retiring baby boom generation. Aggressive RAC attacks are anticipated, and physicians will be targeted for substantial repayments in the coming years.

Will the coming RAC attacks be a disincentive for adoption or EMRs?

Physicians who adopt electronic health records (EHRs) will be facilitating RAC attacks by making more data available for RACs to mine, at low cost, using their proprietary data-mining software. Physicians using EHRs will be in a constant and costly race to upgrade their software so as to comply with ever changing Medicare rules and regulations that make physicians vulnerable to attack.

The end result for physicians will be increasing expense associated with treating Medicare patients in an environment of shrinking payments.

Here are some links from CMS, including a tentative audit schedule by region, a slide presentation about what it means to you and your practice, and some FAQs.