FDA warning on negative pressure wound therapy

Negative pressure (wound vac) therapy was heavily promoted when the technology came out several years ago. It's easy to use and has become very popular. Now we have this warning from the FDA (via Medscape):

Negative pressure wound therapy (NPWT) systems have been associated with 6 deaths and 77 serious complications during the past 2 years, the US Food and Drug Administration (FDA) has announced.

...The FDA points out that NPWT systems are contraindicated for certain wound types (necrotic tissue with eschar present; untreated osteomyelitis; nonenteric and unexplored fistulas; malignancy in the wound; and wounds with exposed vasculature, nerves, anastomotic site, or organs). Several patient risk factors, including being at high risk for bleeding, should be considered before using NPWT.

My guess (and it's only a guess) is that, given the promotion, convenience and ease of use, these devices have been widely employed without proper expertise or attention to evidence.

Canadian Medical Association Journal no longer open access

---as of January. Some content will be free. Here's how it'll work:

Ideally, medical information that saves lives and reduces harm should not be up for sale — at least not to the end-user. Such material should be treated as a public good and paid for by the public, which is why research will remain free of charge. Editorials, news, clinical images, abstracts and previously published articles will also remain accessible to all readers. Access to reviews, analysis, practice, commentaries, humanities and supplements will be restricted to members of CMA, although these items will become free of charge 12 months after publication.

This is disappointing to me, since I have often taken advantage of CMAJ's open access full text content to link as resources here.

BNP and proBNP in the management of heart failure

I've mentioned before the swinging pendulum on this topic, and sure enough, shortly after I wrote this post citing a meta-analysis showing a benefit of guided therapy came the AHA 2009 presentation of TIME – CHF, suggesting NO benefit in terms of 18 month admission free survival. Via AHA 2009—And the Beat Goes On.

Cardiac CT incidentalomas

Although screening for lung cancer with CT is not highly evidence based or widely advocated, many incidental nodules (as well as occasional non-neoplastic incidental abnormalities) are picked up on cardiac CT imaging. A recent paper in JACC reported high cost and low yield, and reached the following conclusions regarding such incidentalomas:

Although noncardiac IF are common, clinically significant or indeterminate IF are less prevalent. Rates of death were similar in patients with and without IF, and IF was not an independent predictor of noncardiac death. The investigation of IF is not without cost or risk. Larger studies are required to assess the potential mortality benefit of identifying IF.

This raises important questions. What should be done with the information? Should radiologists even over-read the non-cardiac portion of cardiac scans as is customarily done?

A fundamental question is: if information is there why let it go down the drain?

Another consideration is driven by anecdotal experience. I am aware of one or two cases in which incidental information was life saving.

Finally, external factors in our culture which drive defensive medicine will always place pressure on the clinician to chase down the lesion. A reasonable compromise might be to let the patient make an informed decision.

Via Heart.org and Heart of the Matter.

Twitter and Facebook no longer free?

Or so it's rumored.

Eric Cropp redux

Since I linked to Bob's post on this subject yesterday he has drawn a couple of commenters. Their remarks were so insightful and informative I thought I'd share portions here.

One commenter provided additional detail about the system problems that prevailed:

...it was a Sunday morning and the HIT devices were experiencing “unplanned downtime”. Directly from the ISMP report:

“The pharmacy computer system was down in the morning, leading to a backlog of physician orders.

The pharmacy was short-staffed on the day of the event

Pharmacy workload did not allow for normal work or meal breaks.

The pharmacy technician assigned to the IV area was planning her wedding on the day of the event and, thus, highly distracted

A nurse called the pharmacy to request the chemotherapy early, so Eric felt rushed to check the solution so it could be dispensed (although, in reality, the chemotherapy was not needed for several hours).”

The commenter went on to question whether down time procedures had been developed or were carried out. In the body of his post Bob mentioned that an empty .9% saline bag on the counter near the chemo mixture suggested that indeed it had been used for the preparation. That raises another question. Was the chemo mixture that was actually delivered labeled as 23% sodium chloride? If so, how many other people, including the patient's nurse, handled the infusion bag? Did they notice? Should they have? If it was not so labeled how (particularly with the computers down) was Cropp to know?

And why the rush to administer the chemotherapy? Was there bed control pressure? Would a time out have helped? Would this have happened back in the pre-HIT days? (Likely not).

Another reader commented on the culture of blame:

Eric’s imprisonment is a tragic demonstration of how early we are in this patient safety journey in the culture/systems of our healthcare organizations, in our understanding of accountability, in medical liability and the courts, and in public policy. Yet I am most worried that we will all fail to learn from this case, fail to ask the question “Could it happen here?” and be writing in a short time about another dead child, another devastated family, and another nurse, doctor, or pharmacist.. Instead of imposing the scarlet letter of accountability on this jailed pharmacist, will we do something about the fact that “Our systems are too complex for merely extraordinary people to perform perfectly 100% of the time?”

Not just yet, apparently.

Eric Cropp: a second victim

Lately, hospitalist leader and patient safety expert Bob Wachter has been making the case for more individual accountability in the the field. But in his Thanksgiving Day post he reminds us that some medical errors, even tragic errors, are indeed system problems.

On February 26, 2006, 2-year-old Emily Jerry died because her chemotherapy was mistakenly mixed with 23% saline rather than normal saline. Eric Cropp, the hospital pharmacist who signed off on the mixture, was the second victim. Despite the fact that a root cause analysis found this to be a case of ordinary human error compounded by system problems the legal system, driven by family outcry, was bound and determined not only to have its pound of flesh but to single out one individual---Eric Cropp.

In addition to suffering criminal penalties (which Emily's mom evidently thought were too lenient) Cropp lost his livelihood. According to one news report Cropp, unable to find steady employment, worked odd jobs cleaning boats and walking dogs.

Wachter sees this as an example of objective analysis trumped by emotions and concludes:

Similarly, it is natural for the parent, spouse or friend of a loved one who dies of a medical mistake to demand a pound of flesh, damn the Swiss cheese. In Internal Bleeding, Kaveh Shojania and I wrote about this, citing the wonderful book on medical malpractice by anesthesiologist Alan Merry and famed novelist Alexander McCall-Smith (yes, that McCall-Smith). The words ring true as we think about the case of Eric Cropp:

“It is to be expected that families or patients will blame the party holding the smoking gun, just as they would a driver who struck their child who ran into the street to get a ball. Some bereaved families… will ultimately move on to a deeper understanding that no one is to blame – that the tragedy is just that. But whether they do or not, write Merry and Smith, ‘It is essential that the law should do so.’”

In this case, the law did not do so. So, as your enjoying your turkey on this day of thanksgiving, please give a thought to the victims of this terrible tragedy, Emily Jerry and her parents.

And Eric Cropp.

A skeptical view of hospital medicine

Only one academic leader I know of has dared challenge the notion (more dogmatic than evidential) that the hospitalist model delivers better and more efficient care. He is Dr. Robert Centor, the blogger at DB's Medical Rants and co-author of a recent editorial on the subject.

“People keep giving me a chance to express my opinion,” says Dr. Centor, who is associate dean at the University of Alabama’s Huntsville Regional Medical Campus and has served as a ward attending for 30 years. “I’m much more willing to say out loud what some hospitalist leaders have tried to keep quiet.”

Read the rest in Today's Hospitalist.

New concepts in status epilepticus

From a review in Emergency Medicine Clinics of North America, here are a few key points:

Definition of SE---

Traditional: 30 minutes of unrelenting seizure(s).

New (proposed): greater than 5 minutes or two or more without intervening full return of consciousness.

Why?

The longer durations traditionally used to define SE were selected based on assumptions about underlying pathophysiology that are now known not to be true rather than on any kind of clinical relevance…

Data from continuous electroencephalographic (EEG) monitoring indicate that the average length of a benign, self-limited, adult generalized tonic-clonic seizure (including pre-tonic-clonic, tonic, and clonic phases) is just longer than a minute and only rarely persists beyond 2 minutes. Patients with seizures that last more than 5 minutes are not likely to improve spontaneously. Consequently, most of these patients are more similar to those with seizures of 30 to 60 minutes' duration than they are to patients with benign seizures of less than 5 minutes' duration. Thus, the new definition is better at discriminating prognostically.

Prehospital treatment by paramedics---effective, safe, reduces likelihood of ICU admission and shows a trend toward reduced mortality.

Initial hospital treatment---

A new accelerated protocol has been proposed in which the previously sequential steps of lorazepam and phenytoin are now stacked and given simultaneously and in which phenobarbital has been omitted. In principle, this accelerated strategy should allow more aggressive treatment of patients who have SE and should reduce the delay to the induction of pharmacologic coma (thereby improving the outcomes) in those with the most refractory seizures. Clinical research to determine if this is true in practice is still required.

The lorazepam dose in the protocol is 0.05-0.1 mg/kg. The phenytoin dose is 18-20 mg/kg with the option to repeat once. Drugs for the induction of pharmacologic coma in refractory SE include midazolam 0.2mg/kg then 1.2 mcg/kg/min. Propofol and pentobarbital are listed as alternatives.

Nonconvulsive SE---need EEG to diagnose.

Alcololic ketoacidisos (AKA)

This is the most recent free full text general topic review I have been able to find.

Key points about AKA:

It is not as benign as popularly believed. It may be a cause of unexplained sudden death in alcoholics.

Acidosis may be severe but if altered LOC occurs suspect another cause.

The treatment is general supportive care and dextrose containing IV fluids (thiamine to go with that, please, and give it first).

The pathophysiology of AKA defies complete understanding. Dr. Daniel Foster, Professor of Endocrinology at UT Southwestern Medical School, believes that AKA is idiosyncratic, with most patients never developing it while a few patients do so repeatedly. A history of recurrent episodes is characteristic. The nature of the idiosyncratic susceptibility is unclear.

Alcohol is metabolized to acetate via sequential steps mediated by alcohol dehydrogenase, and the acetate thus formed can produce acetoacetate. However, the major source of ketone bodies produced in AKA, like DKA, is the catabolic sequence in which free fatty acids (FFA) are released from adipocytes and undergo beta oxidation in the liver mitochondria.

A state of semistarvation (leading to low insulin levels) prevails in patients with AKA in whom alcohol has been their major source of caloric intake. Alcohol intake is often markedly reduced around the time of development of AKA thus accentuating caloric deprivation. If an individual has working beta cells (i.e. does not have DM-1) starvation ketosis is limited and does not progress to ketoacidosis. This is due to a feedback mechanism in which ketones and FFAs feed back on the pancreas, causing it to secrete a small trickle of insulin. In this way, during starvation, the body produces ketones as fuel for the brain (ketones are the brain’s only alternative to glucose) while avoiding fatal ketoacidosis. This regulatory mechanism appears to be ineffective in certain unusual clinical states of which AKA appears to be one. In AKA ketosis continues unfettered. FFAs are much higher than in ordinary starvation ketosis and counter-regulatory hormones are high. Dextrose is therapeutic because it ends the catabolic sequence and shuts off ketosis.

Novel H1N1 influenza in hospitalized patients

JAMA has a recent issue devoted to this topic. Here’s what stands out:

In a series from California admitted patients tended to be younger (27) but of those admitted, those over 50 had the highest mortality (11% overall, 18-20% for those over 50). 66% of chest xrays showed infiltrates and 31% of patients required ICU. Viral pneumonia and ARDS were the most common causes of death. Secondary bacterial infection was only found in 4%.

A Canadian series looked at critically ill patients. Mean age was 32 and 28 day mortality was 14%. Patients’ illnesses progressed rapidly with a mean of 1 day from hospital admission to ICU transfer. Shock and nonpulmonary organ failure were common. Hypoxemia tended to be severe with the mean PO2/FiO2 of 147. 81% got mechanical ventilation with a mean duration of 12 days. Lung rescue therapies were commonly used.

A series from Mexico reported similar outcomes with additional findings of note: 31% were obese and antiviral treatment (adjusting for those with no opportunity for treatment due to early death) improved survival.

Investigators from Australia and New Zealand reported on a series of patients referred for ECMO. Those patients had a 21% mortality by study end. Patients transitioning to ECMO had a median PO2/FiO2 of 56 on a median of 18cm PEEP.

An editorial, commenting on a trial published in the same issue showing non-inferiority of surgical masks compared with N-95s noted that we still don’t have enough evidence on this issue and the controversy will continue.

Recognition of focused practice in HM

Here's more info from The Hospitalist. When I posted about this a couple of months ago I wondered when we would see some resources to help with preparation. In article Scott Flanders said:

“SHM has to help develop the tools and resources hospitalists will need to successfully prepare for and pass this test,” he says. He expects educational resources and self-assessment modules will be available on SHM’s Web site (www.hospitalmedicine.org) and at HM10, April 8-11 in Washington, D.C.

Most hospitalists will work through the RFP process as part of their 10 year recert. What about those who are grandfathered in with lifetime certification? They don't have to take the test but they can. Details on how that will work are unclear.

AHA 2009: chronic pacemaker mediated cardiomyopathy?

We've long known that right ventricular pacing is not physiologic. PACE study investigators, presenting at AHA 2009, looked at the difference between RV and BiV pacing in terms of EF after one year in patients with normal or nearly normal systolic function at baseline.

By one year there was no difference in 6 minute walk performance or quality of life indicators, but patients in the RV pacing group dropped their EF (by 7%) whereas those in the BiV group did not. For a patient starting with a normal EF, a 7% drop may not seem like much, but consider the impact of a few % per year!

RV pacing leads to cardiac remodeling. BiV pacing prevents it. Based on these findings, chronic RV pacing is a bit like depriving some patients of ACE inhibitors.

If you've taken care of many patients with pacemakers you've probably seen pacemaker mediated cardiomyopathy. The hypothesis needs further study before we implant BiV pacers in all patients, but for now there is an important take home message---minimize RV pacing via rate and AV interval programming and be aware that patients in whom this is not possible, because they are RV pacer dependent, may be at special risk to develop dilated cardiomyopathy.

For details on this fascinating study:

Presentation slides.

Discussant slides.

Presentation audio.

Discussant audio.

Summary slide.

Homeopathy is woo. Green Journal.

I love the courage of the Green Journal in calling out quackery. You wouldn't find an article like this in JAMA or the Annals. Of note, one of the authors is identified with a department of complementary medicine. All the same, he knows things that go quack.

Patient safety 10 years after “To Err is Human”

Commentary from key leaders in the latest issue of The Hospitalist reflects on where we've come in the decade since the Institute of Medicine released its patient safety report. Though the article is mainly congratulatory there has been spirited debate about the success (or lack thereof) of the patient safety movement. A recent report by the Consumers Union declared the movement a failure. Although the report lacked data on safety breaches and was based on a faulty premise (equating the continued high rate of hospital associated infections to unsafe practice) it correctly pointed out that we have done little to implement the IOM's recommendations.

Why haven't we done more? Maybe because the IOM's recommendations were predicated on a culture of transparency while over the past decade we have devolved into a culture of blame. These are mutually exclusive outcomes. I pointed that out here and in other posts as my own reason for giving the movement a failing grade.

Meanwhile Bob Wachter gives the patient safety movement less than glowing reviews but apparently for the opposite reason. He seems to believe we have achieved a no blame culture and not only that, it has gone too far. He opened his September 30 blog post with:

In this week’s New England Journal, Peter Pronovost and I make the case for striking a new balance between “no blame” and accountability. Come on folks, it’s time.

Why the seemingly divergent conclusions? Because we've conflated the ideas of accountability and blame. Blame is finger pointing when something goes wrong. The ad below reflects our culture of blame. Note the implied finger pointing: if the patient gets VAP it's someone else's fault.



That's the attitude that pervades medicine today and it's so counterproductive! But it's not the accountability Bob's talking about. If we make careful distinctions between accountability and blame Bob and I are not that far apart. He believes there should be consequences for stubborn, deliberate violation of known safety measures such as hand washing and time outs. I couldn't agree more. It's a form of disruptive behavior, for which Joint Commission already requires “zero tolerance”.

Low blame and high accountability are not conflicting goals. What we have now is too much blame and not enough accountability! In its report the IOM noted that transparency would be essential to effective process improvements for patient safety. The report also stated that such transparency would require movement toward a blame free culture. But, due to misinterpretation of research data and a blunder in the use of language the IOM report had the opposite effect! I explained here:

Ironically, the IOM, with the issuance of its report, undermined its own aspirations for a culture of transparency right out of the gate! They did this by indiscriminately referring to a broad spectrum of adverse patient outcomes as “errors.” This unfortunate attribution, based on faulty analysis of a landmark study on adverse hospital events, I have argued, did much to sabotage the cause of patient safety in the decade that followed. (For the original investigator’s own criticism of the IOM’s interpretation see this editorial). Instead of an era of transparency we entered a heightened culture of blame and finger pointing. The most vivid example of this, of course, is Medicare’s never events policy, of which we are just beginning to realize the adverse consequences.

Now we have a domino effect of more and more unintended consequences. Already in the works is a new legal standard which shifts the burden of proof to hospitals in certain malpractice cases. So much for transparency!

At a recent UWGrand Rounds presentation Bob described what he called a “masterful bit of spin” by the IOM in equating the rate of adverse events in health care to one commercial air crash per day. But the IOM did worse than spin data. They spun assumptions based on faulty interpretation of data. Masterful but unfortunate.

Mitosis

I was taught to remember the acronym IPMAT.

Interphase

Prophase

Metaphase

Anaphase

Telophase

(There’s also preprophase and prometaphase).

Many school kids in the 1960s learned about it from The Thread of Life, one of the Bell Laboratories science films. Clip below.

Novel H1N1 vaccine safety

Monitoring has been intense but there have been no unusual concerns to date according to two reports.

Has public reporting impacted health care at all?

Public reporting of health care performance measures has been a hot topic for several years. It already had a good head of steam in 2005, my first year of blogging, when policy mavens hoped it would be transformative. Last year Bob Wachter even wondered if we were entering a golden age of transparency.

But how was it supposed to work? The conventional wisdom was that it would work by influencing consumers' choices. That never came to pass, though, and Bob suggested another effect:

Shockingly, in the past five years, these mantras have proven to be way off the mark. Instead,

1.Some rudimentary quality data has been placed on the Web
2.Few people are looking at these data, and virtually no real patients are making their healthcare purchasing decisions based on them.
3.And yet… hospitals are doing organizational cartwheels trying to improve their performance on the publicly reported indicators.

Although #2 is surprising, #3 is truly flabbergasting -- it demonstrates the power of shame and embarrassment as motivating forces.

So, shame and embarrassment are motivating hospitals! Yes, it's true, but is it a good thing? In my view such motivation amounts to little more than a form of institutional narcissism---it's all about us (the institution) and our report cards, and has little to do with patients. (As regular readers know I believe that, with few exceptions, performance measures are weak, sometimes non-evidence based and occasionally even harmful).

OK, all cynicism aside: has this transparency movement had any meaningful impact at all? Up to now we've had no evidence one way or the other. That brings me to the EFFECT study, presented at AHA 2009 and published on line in JAMA. The conclusion in the published abstract doesn't give you the real flavor of the study but here's what it said:

Results The publication of the early feedback hospital report card did not result in a significant systemwide improvement in the early feedback group in either the composite AMI process-of-care indicator (absolute change, 1.5%; 95% confidence interval [CI], –2.2% to 5.1%; P=.43) or the composite CHF process-of-care indicator (absolute change, 0.6%; 95% CI, –4.5% to 5.7%; P=.81). During the follow-up period, the mean 30-day AMI mortality rates were 2.5% lower (95% CI, 0.1% to 4.9%; P=.045) in the early feedback group compared with the delayed feedback group. The hospital mortality rates for CHF were not significantly different.

Conclusion Public release of hospital-specific quality indicators did not significantly improve composite process-of-care indicators for AMI or CHF.

So, all in all, public reporting had little impact. I'll get back to that barely statistically significant improvement in MI mortality in a moment.

This study had a somewhat roundabout design. Both comparison groups of hospitals were subject to public reporting. The difference was, in the “intervention group” the reporting was early and accompanied by a media blitz. In the comparison group the reporting was later and there was no media blitz. According to post reporting surveys the intervention hospitals did scramble to enhance their performance, but in a heterogeneous and disorganized manner, accounting for the lack of measured difference in overall adherence to indicators between the comparison groups.

Now what about that improvement in MI mortality? It turns out that it was largely attributable to an improvement in STEMI mortality. Although the percentage of hospitals achieving time-to-reperfusion benchmarks didn't differ between the two groups the findings suggested that in the intervention group there may have been a shorter time to reperfusion. If so it appears to have been driven entirely by time to administration of thrombolytic agents (PCI was used infrequently, but, hey, this was Canada). The only difference in process indicators attributable to the intervention was the frequency of thrombolytics administered in the ER, before transfer to intensive care. Of note, the survey indicated that ten hospitals in the intervention group decided, after the reporting, to allow ER physicians to administer thrombolytics without specialty consultation.

Although I believe performance measures tend to be weak, time to reperfusion is a notable exception. The process change by which hospitals gave ER docs autonomy in the administration of thrombolytics may have made a difference.

So, after all these years and all this momentum, evidence for an impact of public reporting is slim. Institutional narcissism, up to now, has been largely ineffective, being driven by weak and perfunctory process measures. The quality movement needs to adopt a more thoughtful and nuanced approach to improvement.

AHA 09: the POPE study

Hospitalists involved in the care of CABG patients will find this study of interest. Pericardial irritation occurs post CABG. Pericarditis can cause arrhythmias as well as lead to effusion progressing to tampanade. Frequently NSAIDs are given to reduce these complications. Evidence to support this strategy has been lacking. The POPE study, presented at AHA 09, looked at the use of diclofenac to reduce the volume of pericardial effusion. The findings, from the press release:

A non-steroidal anti-inflammatory drug, diclofenac, failed to reduce fluid accumulation around the heart after cardiac surgery.

The NSAID after heart surgery also did not reduce the serious problem of cardiac tamponade, compression of the heart by fluid.

Only patients with pericardial effusion at baseline were enrolled. The investigators concluded:

The use of NSAIDs in post-cardiac surgery patients is useless in this setting.

As pointed out in the presentation slides NSAID use post CABG has been common even though not supported by good evidence, and in spite of known adverse effects including increased renal failure, heart failure and myocardial infarction.

Amyloid cardiomyopathy

This Clinical Perspective piece in Circulation stresses the little appreciated fact that there are actually three types of cardiac amyloidosis: the AL type that we’re used to thinking of as well as two types of transthyretin related disease, the mutant type (of which there are over 100 mutations) and the wild type seen almost exclusively in older men. The three types differ in clinical features and prognosis.

EMRs don't improve health care---yet

Claims made over the last several years that EMRs lead to better quality of care have been devoid of supporting evidence. Now we have evidence from a large study that EMRs in fact don't improve quality. I apologize for not having a better source for this than a New York Times article, which reports that the findings were to have been reported at a conference in Boston. Unfortunately, the article didn't provide a link to the conference proceedings and didn't name the conference. Maybe I'll find the primary source later. In the mean time the article reports:

But a new study comparing 3,000 hospitals at various stages in the adoption of computerized health records has found little difference in the cost and quality of care.

“The way electronic medical records are used now has not yet had a real impact on the quality or cost of health care,” said Dr. Ashish K. Jha, an assistant professor at the Harvard School of Public Health, who led the research project.

Unfortunately the study looked at performance measures, and performance measures do not equal quality. But the study still makes a strong statement because performance measures represent the bare minimum standards, the lowest of the low hanging fruit. So if EMRs didn't improve performance measures you can be darn sure they had little or no impact on real quality.

I agree with the comments of the interviewees in the article, that we have years to go before we utilize EMRs effectively, in a way that will improve patient care.

Autoimmune polyendocrine syndromes

Reviewed here (Expert Review of Endocrinology and Metabolism).

Colchicine toxicity

Brief review and lessons from the courtroom.

AHA 09: the FOCUS study

This study addressed an issue that often plagues hospitalists who are involved in co-management of patients undergoing hip fracture repair: when the patient's hemoglobin and hematocrit drop following surgery, what is the appropriate transfusion target in patients with cardiovascular disease or risk factors for CVD? While a H/H target of 8/24 has been validated for patients across the board, conventional wisdom has it that 10/30 is a more appropriate target for patients with cardiovascular disease. FOCUS compared strategies utilizing these two targets and found no advantage of the more aggressive transfusion strategy in terms of death or cardiac event rate. Results concerning the effect on functional outcomes have yet to be announced.

From AHA 09 scientific sessions: CT-STAT

It's always fun, after national meetings, to observe how the popular press botches the reporting of important findings that were presented. CT-STAT compared two diagnostic strategies for ER evaluation of chest pain in low and intermediate risk patients: initial cardiac markers and electrocardiograms followed by either traditional stress imaging or CT coronary angiography. Here's an Associated Press report on the findings. Nowhere in the report is there a link to the primary source (other than a link to the AHA website) and nowhere is the name of the study mentioned. Moreover, the article failes to inform readers of which of the several types of CT scans was used (it was CT coronary angiography). The article gets it wrong right out of the gate:

A CT scan — a kind of super X-ray — provides a faster, cheaper way to diagnose a heart attack when someone goes to the emergency room with chest pains, a new study suggests.

On the whole CT scanning was indeed cheaper and resulted in faster determination of disposition. And, it was just as good as traditional work up in terms of freedom from events, short term and at 6 months. What's wrong with the statement is the notion that CT diagnosed heart attacks. Wrong, wrong, wrong. Patients enrolled in CT-STAT had already ruled out for MI via cardiac biomarkers and electrocardiograms. CTA is a rule out, not a rule in, test. Over two thirds of patients undergoing CTA had normal or near normal coronary arteries. Of the remaining patients it was not always possible to determine whether visualized disease was significant or “culprit” in nature, and this led to further evaluation.

An important caveat is that the CT strategy is not applicable to higher risk patients, in whom the negative predictive power is not as good. For those patients who rule in subsequent evaluation may be complicated. First, a cardiac catheterization would expose the patient to a second dye load and increased risk of renal failure. This might necessitate waiting an extra day with hydration before proceeding to cath. Second, after visualization of disease on CTA functional testing may then be necessary to establish the physiologic significance of the lesion(s). In other words, some patients, after CTA, may end up getting a stress nuc anyway. For patients who don't rule out with CTA the subsequent diagnostic strategies may in fact be slower and more expensive. We really don't know because the study was not designed to answer that question. Based on these study results, though, the CT strategy is cheaper and faster over all when used in low risk patients because the cost and time savings in the patients who rule out outweigh any increase in cost for the other patients, who are in the minority.

Additional resources about this study, direct from AHA, can be found here, here and here.

ARBITER 6–HALTS---what does it really mean?

This is somewhat old news by now, but its results have been widely distorted in popular media. With this post I hope to put the issue in perspective.

From the NEJM abstract (italicized comments mine):

Methods We enrolled patients who had coronary heart disease or a coronary heart disease risk equivalent, who were receiving long-term statin therapy, and in whom an LDL cholesterol level under 100 mg per deciliter (2.6 mmol per liter) and an HDL cholesterol level under 50 mg per deciliter for men or 55 mg per deciliter for women (1.3 or 1.4 mmol per liter, respectively) had been achieved. (Since when is an HDL cholesterol below a certain target an achievement?). The patients were randomly assigned to receive extended-release niacin (target dose, 2000 mg per day) or ezetimibe (10 mg per day). (Added to statin therapy). The primary end point was the between-group difference in the change from baseline in the mean common carotid intima–media thickness after 14 months....

Results The mean HDL cholesterol level in the niacin group increased by 18.4% over the 14-month study period, to 50 mg per deciliter (P less than 0.001), and the mean LDL cholesterol level in the ezetimibe group decreased by 19.2%, to 66 mg per deciliter (1.7 mmol per liter) (P less than 0.001). Niacin therapy significantly reduced LDL cholesterol and triglyceride levels; ezetimibe reduced the HDL cholesterol and triglyceride levels. As compared with ezetimibe, niacin had greater efficacy regarding the change in mean carotid intima–media thickness over 14 months (P=0.003), leading to significant reduction of both mean (P=0.001) and maximal carotid intima–media thickness (P less than 0.001 for all comparisons). Paradoxically, greater reductions in the LDL cholesterol level in association with ezetimibe were significantly associated with an increase in the carotid intima–media thickness (R=–0.31, P less than 0.001). The incidence of major cardiovascular events was lower in the niacin group than in the ezetimibe group (1% vs. 5%, P=0.04 by the chi-square test).

From the conclusion:

...extended-release niacin causes a significant regression of carotid intima–media thickness when combined with a statin and that niacin is superior to ezetimibe.

Here are the take home points:

This trial does not refute the LDL hypothesis. For decades diet and drug trials have shown that the more LDL is lowered the fewer the events. The evidence to support this relationship is reproducible over time and is overwhelming.

It makes intuitive sense that lowering of cholesterol by an absorption blocker would be beneficial. An older class of absorption blockers, the bile acid resin binding agents, such as cholestyramine, has been shown to reduce cardiovascular events. Their mechanism, however, inhibition of reabsorption of bile acids thereby consuming cholesterol to make more bile acids, differs from that of ezetimibe. Ezetimibe directly inhibits cholesterol absorption by a novel mechanism as explained here in the product labeling:

The molecular target of ezetimibe has been shown to be the sterol transporter, Niemann-Pick C1-Like 1 (NPC1L1), which is involved in the intestinal uptake of cholesterol and phytosterols.

Evidently, though, ezetimibe does more, and that was the new finding in this study: it lowered HDL cholesterol. HDLC, contrary to popular distortion, is not “good cholesterol” that protects against atherosclerosis. Instead, the concentration of HDLC is a marker of the efficiency of reverse cholesterol transport. Ezetimibe is absorbed systemically and who knows what it's doing outside the intestinal brush border. So, if the finding of this study is real, that ezetimibe really does lower HDLC, that means it may be doing something to impede reverse cholesterol transport. That would be bad indeed. Niacin, on the other hand, has been known for years to improve reverse cholesterol transport, thereby raising HDLC.

The finding that ezetimibe was associated with a lowering of HDLC is indeed surprising. The final study conclusions, though, are not surprising for this reason: it has been well understood for decades that, despite the clinical efficacy of lowering LDL, it's not enough in many patients. Consider all those statin trials that show relative risk reductions on the order of 30%. Laudable as that is it's far from 100%! Thus was born the notion of residual risk after statin therapy. Once LDLC is lowered to goal, via statin therapy or other means, many patients remain at risk. Niacin addresses that residual risk. Further LDLC reduction does not. Why?

The residual risk is composed of a mixed bag of metabolic disturbances, most of them attributable to the metabolic syndrome which includes impaired reverse cholesterol transport, insulin resistance, elevation of triglyceride-rich atherogenic particles and small LDL particle diameter. Niacin reverses these disturbances. For an in depth explanation of these risk factors check these references: [1] [2] [3].

One finding was reassuring. Combo therapy with niacin and a statin was safe. Although it was a small study there were apparently no issues with increased myositis or hepatotoxicity risk attributable to the combination as is seen with statin-fibrate combinations.

How should hospitalists be cultivated?

Here are the two competing models: 1) as clinicians at the top of their game and 2) as stage performers. I’ve been in the field since its beginnings and over the years I’ve seen it devolve from model number 1 towards model number 2. If the practice management track at SHM 2009 is any indication, model 2 is winning. Not a good thing, in my opinion.

The title of hospitalist leader Mary Jo Gorman’s talk about running a program is a dead giveaway: “Performance Reviews: Dealing with the Problem Performer.” I don’t run a hospitalist program but if I did I’d be more interested in how my hospitalists functioned as clinicians than how they performed. Anyway, the focus of the talk was on behavior and conformity. Clinical skills were mentioned only in passing.

There were some good points in the talk. Most importantly, don’t expect the hospitalists on your team to read minds. Be clear in your expectations. Good, explicit communication (not just abstract goals like “be a team player”) solves a lot of problems. And yes, rules are important, but tools---tools that help the hospitalist function optimally as a clinician---are even better.

Monitoring patients for HIT

Guidelines exist for monitoring hospitalized patients for the development of HIT and for evaluation of patients who develop thrombocytopenia. How well are they followed? Not very well at one academic medical center in the Netherlands:

The frequency of compliance with platelet count monitoring recommendations was 26.3% for all patients receiving dalteparin, 35.6% for all patients receiving nadroparin, 23.0% for surgical patients receiving prophylactic dosing of either dalteparin or nadroparin, and 41.5% for patients exposed to UFH within 100 days before the start of either dalteparin or nadroparin treatment. Regular platelet count monitoring was strongly positively associated with medical patients (relative risk [RR] 2.33), surgical patients (RR 2.03), critically ill patients (RR 2.80), and those with recent exposure to UFH (RR 2.19). The frequency of testing for HPF4-Ab was 5.4% and the initiation of alternative anticoagulation with danaparoid in patients with a 50% drop in platelet count was 0%.

Proper diet and exercise may stem cognitive decline

Via Internal Medicine News.

Another review of autoimmune pancreatitis

Via Expert Review of Gastroenterology and Hepatology.

Key points:

This is a newly appreciated clinical entity.

It is a mimic of pancreatic carcinoma.

It is very responsive to corticosteroids.

Elevated IgG4 is characteristic.

Mechanical ventilation 101---understanding the modes

This review in CCJM can be considered a companion paper to an earlier one I linked here, explaining the terminology of modes of mechanical ventilation. The CCJM review bridges the gap between theory and practice, focusing more on clinical application. It briefly covers basic terminology and focuses on newer (“alternative”) modes:

We call these innovations “alternative” modes to differentiate them from the plain volume-control and pressure-control modes. Some clinicians rarely use these new modes, but in some medical centers they have become the most common ones used, or are being used unknowingly (the operator misunderstands the mode name).

Say, for example, your hospital uses Dräger ventilators. You order CMV, and dial in a tidal volume. Think your patient's getting simple volume controlled ventilation? Think again.

Methadone and Torsades

This review contains a summary of the literature and recommendations for safety monitoring.

The ECG that screams for bicarb

When AVR points toward heaven in the appropriate clinical setting your patient may be headed there before her appointed time without prompt action on your part. In another of his entertaining and fascinating discussions on telltale ECG patterns which guide emergency decision making Dr. James Roberts discusses the electrocardiographic toxidrome of tricyclic antidepressant poisoning.

ECG changes in STEMI

This review in the American Journal of Emergency Medicine describes in detail the sequence of electrocardiographic changes over time in STEMI.

Stress testing in low risk chest pain patients---is it always necessary?

Stress testing for patients presenting to the hospital with chest pain, particularly in this era of order sets and templates, has become knee-jerk. For low risk patients under 40 it may not be necessary according to this study:

In our study, a combination of age younger than 40 years, nondiagnostic ECG result, and 2 sets of negative cardiac biomarker results at least 6 hours apart identified a patient group with a very low rate of true-positive provocative testing. Routine stress testing added little to the diagnostic evaluation of this patient group and was falsely positive in all patients who consented to diagnostic coronary angiography (4 of 6 cases).

If nothing else, maybe stress testing could be deferred to out patient testing in some low risk patients who “rule out”, resulting in considerable cost savings to the hospital.

H/T to The Hospitalist.

DDAVP for prevention and treatment of overcorrection of hyponatremia

Osmotic demyelination is a much feared complication of treatment of hyponatremia. Slow and limited correction minimizes the risk and recommended rates of correction have been widely published. In spite of this, overcorrection is common. This occurs inadvertently because renal diluting capacity is often restored during treatment. A frequent example is treatment of hypovolemic hyponatremia with isotonic saline in which case volume expansion eliminates the signal for vasopressin secretion, resulting in a water diuresis. These changes are unpredictable, calculations of sodium deficits are unreliable, and the rise in serum sodium is often more rapid than anticipated or desired.

These issues were discussed, with a focus on the use of DDAVP, in a paper from last year in the Clinical Journal of the American Society of Nephrology.

I first read about this use of DDAVP several years ago in UptoDate but could find little to support it in the primary literature. The paper referenced here is the largest published experience that I'm aware of. It is based on a retrospective chart review and its findings and conclusions are as follows:

Results: Six patients (group 1) were given desmopressin acetate after the 24-h limit of 12 mmol/L had already been reached or exceeded; correction was prevented from exceeding the 48-h limit of 18 mmol/L in five of the six. Fourteen patients (group 2) were given desmopressin acetate in anticipation of overcorrection after the plasma sodium concentration had increased by 1 to 12 mmol/L. In all 14 patients who were treated with desmopressin acetate as a preventive measure, correction was prevented from exceeding either the 24- or 48-h limits. After desmopressin acetate was administered, the plasma sodium concentration of 14 of the 20 patients fell by 2 to 9 mmol/L. In all six group 1 patients and in five of the group 2 patients, the plasma sodium concentration was actively lowered again by the concurrent administration of desmopressin acetate and 5% dextrose in water; no serious adverse consequences from this maneuver were observed.

Conclusion: Desmopressin acetate is effective in preventing and reversing inadvertent overcorrection of hyponatremia.

The link above is to the entire paper free of access controls, which is worth reading in its entirety. I was struck be several points:

The authors are experienced clinicians who are thoroughly familiar with the pathophysiologic rationale for the use of DDAVP. When used by such clinicians DDAVP is safe and effective in preventing and mitigating overcorrection of hyponatremia. The authors present several pearls governing effective use.

Impending overcorrection is often signaled early by a return of renal diluting capacity evidenced by urine volume, urine osmolality, or the initial rate of rise in sodium concentration.

Hyponatremia in this series was often multifactorial but the frequency with which thiazides and SSRIs were involved was impressive.

Clinical scenarios of inadvertent overcorrection include volume expansion, discontinuation of thiazides and SSRIs, and treatment of adrenal crisis with glucocorticoids.

Perioperative care of patients with obstructive sleep apnea

One of the speakers at SHM 09 mentioned that there were guidelines out concerning this. I was not aware they existed. They make some fairly strong statements which hospitalists need to be aware of, so I'm posting them here.

Atherosclerotic renal artery stenosis

This review covers the ins and outs of medical treatment including secondary prevention of atherosclerotic events and the safety of ACEIs. Concerning stenting:

In conclusion, the additional value of renal artery stenting needs to be established in patients who are aggressively medically managed; and a functional test to help us discern patients who would benefit from renal revascularization from patients who would not is needed. Because the issue of benefit from renal artery revascularization is still undecided, patients should be encouraged to enroll in randomized clinical trials. Only when we have solid evidence that patients will benefit significantly from renal artery revascularization in addition to optimal medical therapy should we incorporate RAS revascularization into routine medical practice.

Systolic and diastolic heart failure represent a false dichotomy

---one which was forced on us by clinical trial design and perpetuated by administrative documentation requirements.

Read about it here.

Steroids for meningitis in adults and adolescents

What's the latest evidence?

According to a recent meta-analysis from Mayo Clinic Proceedings:

The administration of corticosteroids resulted in a lower short-term mortality rate than did the administration of placebo in high-income countries (pooled RR, 0.5; 95% CI, 0.27-0.92; I2=0%) and in the studies with a low prevalence of infection with HIV (RR, 0.66; 95% CI, 0.44-0.99; I2=0%). In studies from high-income countries, the number needed to treat with corticosteriods to prevent 1 death and 1 neurologic sequela was 12.5 (95% CI, 7.1-100.0) and 11.0 (95% CI, 5.6-100.0), respectively.

CONCLUSION: Our meta-analysis suggests that the adjunctive administration of corticosteroids is beneficial in the treatment of adolescents and adults with bacterial meningitis in patient populations similar to those seen in high-income countries and in areas with a low prevalence of HIV infection.

Non-acetaminophen induced acute liver failure---is there a role for N-acetylcysteine?

I've been hearing some noise about this lately. What does the evidence show? I got six citations doing a PubMed search with the following strategy:

("acetylcysteine"[MeSH Terms] OR "acetylcysteine"[All Fields] OR "n acetyl cysteine"[All Fields]) AND non-acetaminophen[All Fields]

Here they are:

This small study compared treated patients with a historic controls and demonstrated reduced mortality. Oral NAC was used.

This recently published RCT used IV NAC and demonstrated improved transplant free survival if used early. (For patients with advanced encephalopathy nothing short of transplant seems to salvage them).

This Chest review of ALF makes mention of it as a possible modality in non-acetaminophen ALF.

This paper suggests efficacy and safety in the pediatric population.

This systematic review was negative but out of date, with no citations after 2003 included.

One other paper examined pathyphysiologic effects in mouse liver.

All in all the evidence for benefit is mounting and the treatment appears safe. Not standard of care yet but maybe worth a try.

More evidence that guideline adherence improves outcomes

These data are from the Get with the Guidelines program, for heart failure and MI.

Endemic mycoses

We're talking, of course, about histo, blasto and cocci.

Here are a couple of pearls from a review in Clinics in Chest Medicine:

All can present as “CAP.” Coccidioidomycosis accounted for 29% of CAP cases in one Arizona survey!

Endemic mycoses, particularly blasto, can present as ARDS.

A systematic approach to diagnosis is outlined.

Hyponatremia in hospitalized patients

This is one of many reviews on this topic but it's not just any review. It is the most thorough examination of the pathopysiology I've seen in a long time. Read it when you're well rested and well fed. Two or three times.

Via Medscape. CME available.

Heparin treatment of extreme hypertriglyceridemia

Extreme hypertriglyceridemia is a significant threat when it occurs in hospitalized patients. It typically presents itself as a call from the lab that the patient's serum is lipemic. The patient is at risk for acute pancreatitis if s/he doesn't already have it. Compounding the management difficulty is the interference by lipemia with certain laboratory procedures, the best example of which is pseudohyponatremia although there are also other laboratory problems, and special analytical methods are needed.

The Archives of Internal Medicine presents a case report and literature review of the use of heparin to rapidly clear triglycerides. The patient in question had DKA. Because lipemia is often post prandial NPO status itself may reduce triglycerides to a safe level. Insulin is a cofactor for lipoprotein lipase and its use often results in rapid triglyceride clearance. This patient proved refractory, and low dose intravenous heparin added to the regimen was effective. Heparin promotes release of lipoprotein lipase from endothelial cells

Like this paper, most of the literature concerning heparin for the treatment of extreme hypertriglyceridemia is anecdotal or composed of small case series. The reported effects, however, have been dramatic and the effect has been known for decades. Drugs such as fibrates and fish oil are indicated long term but their action may be too slow for the acute situation.

Here are some papers on heparin treatment of hypertriglyceridemic pancratitis: [1] [2] [3] [4].

Statins and COPD outcomes

Statin treatment is the first thing to come along since oxygen and smoking cessation to influence the long term outcome in COPD patients. From a new systematic review in BMC Pulmonary Medicine:

Outcomes associated with treatment with statins included decreased all-cause mortality in three out of four studies (OR/HR 0.48–0.67 in three studies, OR 0.99 in one study), decreased COPD-related mortality (OR 0.19–0.29), reduction in incidence of respiratory-related urgent care (OR 0.74), fewer COPD exacerbations (OR 0.43), fewer intubations for COPD exacerbations (OR 0.1) and attenuated decline in pulmonary function. The RCT reported improvement in exercise capacity and dyspnea after exercise associated with decreased levels of C-reactive protein and Interleukin-6 in statin users, but no improvement of lung function.

Medscape CME here.

Should family members be present during CPR?

Although “open resuscitation” is increasingly popular what is the impact on resuscitation performance? From a study published in Critical Care Medicine:

The time to critical events was similar across groups with respect to initiating CPR, attempting to intubate the patient, and pronouncing the death of the patient. However, the time to deliver the first defibrillation shock was longer for the overt reaction witness group (2.57 minutes) as compared with the quiet (1.77 minutes) and no family witness (1.67 minutes) groups. Additionally, fewer total shocks were delivered in the overt reaction witness groups (4.0 minutes) vs. the quiet (6.5 minutes) and no family witness groups (6.0 minutes).

Conclusion: The presence of a family witness may have a significant impact on physicians' ability to perform critical actions during simulated medical resuscitations. Further study is necessary to see if this effect crosses over into real clinical practice and if training ameliorates this effect.

H/T to The Hospitalist

The obvious limitation here is that the study was of simulated resuscitations.

When patients admitted with COPD exacerbation die in the hospital

---it's more often due to co-morbid conditions than from the COPD itself:

Forty-three patients with a hospital admission diagnosis of COPD exacerbation underwent autopsy; all had died within 24 h of admission to the hospital. Twenty-three patients (54%) had a long COPD history (greater than 10 years), and 19 patients (44%) had more than one hospitalization in the last year of life. The median age at death was 70 years (interquartile range, 65 to 75 years), and male sex was predominant (n = 31; 72%). The main (primary) causes of death were reported as cardiac failure (n = 16; 37.2%), pneumonia (n = 12; 27.9%), and pulmonary thromboembolism (PTE) (n = 9; 20.9%). Respiratory failure due to a progression of COPD was the primary cause of death in six patients (14%). Most patients had more then one comorbid disease (n = 33; 77%), and the most frequent comorbid disease was chronic heart failure (n = 25; 58%).

Eroding empathy

Med students were monitored over time via a physician empathy scale. The findings:

Statistical analyses showed that empathy scores did not change significantly during the first two years of medical school. However, a significant decline in empathy scores was observed at the end of the third year which persisted until graduation....

Conclusions: It is concluded that a significant decline in empathy occurs during the third year of medical school. It is ironic that the erosion of empathy occurs during a time when the curriculum is shifting toward patient-care activities; this is when empathy is most essential.

Ironic but intuitive.

When patients present with their first seizure

Who needs AEDs and which ones? What work up?

Here's a concise discussion of the ins and outs.

Therapeutic hypothermia after cardiac arrest

Beneficial in the latest Cochrane review.

Medscape Heartwire coverage here.

Ischemic colitis for hospitalists

Very helpful nuts-and-bolts review from CCJM.

Balloon pumps for the rest of us

Here's a primer on intra-aortic balloon counterpulsation. Full text from Medscape.

Hyponatremia, gait disturbance and patient falls

This article is a little dated. I just ran across it and found it of interest. Mild, chronic hyponatremia in the elderly is common and often thought best left alone. In this study it was associated with gait disturbance and fall risk.

More acid base resources

I'm linking to this Wikipedia article for the links at the bottom. Some of them are dead, but there are several excellent sites on the list.

The Acid Base Book

---is an on line, open access textbook on acid base disorders currently under development as a labor of love by Robert Centor, MD, the blogger at DB's Medical Rants. Although it deals with some advanced concepts in acid base disorders it is case based and written in plain language.

Comments are enabled; the site exemplifies the beauty of Web 2.0. This is a wonderful resource. More chapters are planned, but if you can't wait check out additional case discussions in DB's Medscape cases.

NHLBI CPR study halted

The ROC PRIMED trial, comparing different EMS protocols for cardiac arrest, was halted after data analysis showed no difference in outcomes. The two arms of the trial evaluated 1) an impedance threshold device attached to the airway and 2) two different durations of CPR applied before rhythm assessment and (when indicated) defibrillation. The NIH press release is a good example of the obfuscation that results in many attempts to make scientific findings palatable to the lay public. It reads in part (my emphasis):

Enrollment has ended early in a large, multicenter clinical trial comparing two distinct resuscitation strategies delivered by emergency medical service (EMS) providers to increase blood flow during cardiac arrest. The study's independent monitoring board and the National Heart, Lung, and Blood Institute (NHLBI), the lead sponsor of the study, stopped enrollment based on preliminary data suggesting that neither strategy significantly improved survival. One strategy compared different durations of manual cardiopulmonary resuscitation (CPR) by EMS providers before they assessed whether defibrillation was needed, and the other strategy tested the potential benefits and risks of an investigational device to maintain pressure in the chest during CPR.

The italicized portion of the above paragraph is deceptive. It omits background information from prior research suggesting that three minutes of compressions applied before assessment for defibrillation (as compared to no compressions) dramatically improves survival. (Background here, here and here). What was not known before was the optimal duration of compressions before defibrillation assessment. After these results, we know, as we did before, that for unwitnessed arrest compressions before assessment for defibrillation are beneficial, and we still don’t know the optimal duration.

The press release goes on, several paragraphs down:

"The ROC PRIMED study answers a long-standing question in the EMS community over whether it is better to defibrillate earlier or later when trying to resuscitate a patient," said Ian Stiell, M.D., professor and chair of the Department of Emergency Medicine at the University of Ottawa, senior scientist at the Ottawa Hospital Research Institute, and a principal investigator for the ROC PRIMED Analyze Early vs. Later protocol. "Both techniques appear to be equally beneficial."

Again, because it omits important background, the statement is confusing and deceptive. If the patient is in the electrical phase of resuscitation early is clearly better. If in the hemodynamic phase later is better---we just don’t know how much later. (For background on the importance of the physiologic phases of resuscitation consult this review).

The Steering Committee chair said this (my emphasis):

Myron Weisfeldt, M.D., ROC Steering Committee chair and director of the Department of Medicine at the Johns Hopkins University School of Medicine in Baltimore, added, "Questions like this one — which address the relative benefits of current medical practices — are an important example of comparative effectiveness research and, in this case, can help advance emergency medical care."

The press release does a poor job of reporting the primary research design and conditions but we have this clue:

The other principal strategy studied in ROC PRIMED was the timing of assessing the heart's rhythm to determine whether defibrillation is needed in relation to when CPR is started. For patients randomly assigned to the Analyze Early group, EMS providers were instructed to perform CPR until they were able to analyze the patient's heart rhythm (approximately 30 to 90 seconds). Patients in the Analyze Later group received CPR for at least three minutes before their heart rhythm was analyzed. When indicated, defibrillation was provided.

We’ll have to wait for official publication of the results, but it appears from this that there was considerable time variability in both arms. Some patients, we don’t know how many, actually got a minute and a half of compressions in the “short” arm. Patients in the long arm received “at least” three minutes. There’s probably an upper limit of optimal time. How long is too long? These questions illustrate a limitation of comparative effectiveness research (CER): experimental designs are inherently more complex than simple RCTs, and more variables are introduced. It’s anybody’s guess what was actually being compared in this study. It may be that the exact time is not critical, there being a range of optimal time. Equally plausible is the possibility that this study compared “too long” with “not long enough.” Because the design tweaks and time combinations that could be studied are virtually endless, CER is unlikely to give us a better answer than we already have from the work of Ewy and colleagues. What the CER agenda is likely to do is delay even longer the much needed guideline implementation of this important work. Medscape commentary here.