Monday, October 31, 2011

Xigris: end of story?

Just over a week ago I posted an update on Xigris, concluding that the results of PROWESS-SHOCK would give us answers to help resolve some of the controversy surrounding the drug. Little did I know that the results would be announced so soon. For those who haven't heard, there was no significant difference attributable to Xigris, neither benefit nor harm, so Lilly has pulled it from the market.

According to this Medscape piece and other reports I've read the mortality in PROWESS-SHOCK was similar to that of the Xigris treatment arms of PROWESS and other Xigris trials. That means it was lower than the placebo arm of PROWESS. Impressive considering that PROWESS-SHOCK was supposed to represent the sickest of patients with sepsis. It may be that in the decade since PROWESS we've gotten better in other areas of sepsis treatment. Early goal directed therapy is gradually catching on, we've refined our approach to the use of steroids and glycemic control and perhaps we are more intelligent in the use of antibiotics. Awareness of the early diagnosis of sepsis has increased. It's ironic that Lilly's support of the Surviving Sepsis Campaign, which improved all these areas of treatment, may have diminished the role of Xigris.

No doubt the media and the pharmascolds will spin the results to read “Xigris doesn't work.” Simplistic. Xigris was a lifesaving drug in the right patient at the right time. Unfortunately that patient was just too hard to define. So, I think Lilly made the right decision to pull the drug.

Many so called NSTEMIs are really STEMI equivalents

That is, they represent acute epicardial coronary occlusion nonetheless. These tricky situations have deprived many patients of a timely trip to the cath lab, even though the infarcts can be just as serious as STEMIs meeting the generally accepted criteria. Many of these infarcts are termed electrocardiographically silent although in most cases, under the careful scrutiny of an experienced reader subtle abnormalities come to light.

Dr. Smith at his ECG blog presents such a case here. This was a Cx/OM occlusion with subtle ECG changes but ultimately a very large infarction. Aside from the fact that the circumflex territory is relatively electrocardiographically silent this was probably a posterior infarction, which characteristically lacks classic STEMI changes.

I have seen countless examples of this sort of thing between my own practice and numerous postings on ECG blogs. These represent patients with STEMI equivalents whose door to balloon times are prolonged due to the subtlety of their ECG manifestations. These long door to balloon times are not tracked as a quality measure because the cases are not captured by the simplistic criteria (ST segment elevation of at least 1 mm in two contiguous leads) for the performance metric.

Despite progressive improvement in door to balloon time as a performance measure overall mortality from acute MI has not declined over the last several years. This may reflect the fact that the performance measurement represents only a fraction of the patients who present with acute coronary occlusion.

Cases like this also help us understand the recent findings in this paper, that STEMIs and NSTEMIs are indistinguishable in terms of prognosis, clinical course and pathology, leading the author, Brendan Phibbs, to conclude that STEMI vs NSTEMI is a baseless distinction. ST segment elevation will identify many patients who need an immediate trip to the cath lab, but simple reliance on this finding will result in some patients with acute coronary occlusion being deprived of timely reperfusion.

There is no obvious solution to this problem. Lack of typical ECG changes in a patient with acute coronary occlusion puts both patient and clinician at a disadvantage. Dr. Smith's advice is that if chest pain is unremitting, then as soon as one has evidence that the pain is ischemic an urgent trip to the cath lab is warranted, ST elevation or not.

Friday, October 28, 2011

Glucocorticoid-Remediable Aldosteronism

This entity was recently featured at Renal Fellow Network. From a review in Cardiology in Review:

Glucocorticoid remediable aldosteronism (GRA) appears to be the most common monogenic form of human hypertension. As a result of chimeric gene duplication, aldosterone is ectopically synthesized in the zona fasciculata of the adrenal gland under the control of adrenocorticotropin (ACTH). Affected individuals are typically hypertensive, often with onset in youth, and demonstrate refractoriness to standard antihypertensives such as angiotensin-converting enzyme inhibitors and β-blockers. GRA subjects are normokalemic but often develop hypokalemia when treated with a potassium-wasting diuretic. Analysis of affected kindreds has demonstrated a high prevalence of early cerebral hemorrhage, largely as a result of aneurysms. Identification of affected individuals should allow direct neurovascular screening and targeted antihypertensive therapy.

The treatment is with steroids to suppress ACTH.

Oncologic emergencies

A two part series from Emergency Medicine. Free full text.

The hazards of NSAIDs---why doesn't the FDA get it?

A great post at The Kidney Doctor talks about NSAID risks and asks why doctors don't get it. It summarizes the adverse renal effects of NSAIDs and links to an article documenting the degree of NSAID use among CKD patients, both over the counter and by prescription.

I would take the discussion a step further and ask why the FDA doesn't get it. Their inconsistency is remarkable. After all for more harm is attributable to NSAIDs than to poor old Darvocet.

The post describes long term NSAID use in CKD patients. But let's not forget the harm caused by only occasional use of NSAIDs in healthy individuals.

Thursday, October 27, 2011

Drug induced ANCA disease

Via Renal Fellow Network. I had heard this mentioned at a talk but had not seen much written about it.

Another review of ITP

---with a focus on thrombopoietin receptor agonists. NEJM review here, commentary in the NEJM blog here.

DRESS syndrome: another review

This one in the American Journal of Medicine.

Recent antibiotic exposure, microbiology and outcomes in severe sepsis

In this study antibiotic exposure in the prior 90 days was associated with increased resistance and increased mortality. The authors recommend taking into account recent antibiotic history in formulating an empiric regimen for sepsis.

Wednesday, October 26, 2011

Pulmonary manifestations of the rheumatic diseases

An entire issue of Clinics in Chest Medicine last year was devoted to this topic. Although much overlap exists, specific pulmonary syndromes are characteristic of specific diseases. Here is a selection of some of the more typical examples. This list is not comprehensive and access to the full text of the entire issue is recommended. Although infections (particularly opportunistic) and pulmonary toxicities of anti-rheumatic drugs are not emphasized here they are important in the differential diagnosis.

Systemic sclerosis (SSc)

Interstitial lung disease---various histologic patterns.

Pulmonary hypertension (PAH)---believed to be more common in lcSSc (formerly CREST) than in dcSSc (formerly PSS) but this belief is in dispute.

Pleural disease---fibrosis, may be clinically silent. Effusions rare.

Chronic aspiration of gastric contents (with esophageal dysmotility).

Pneumonia risk (with myositis overlap).

Mixed connective tissue disease (MCTD)

Though often confused with terms like “overlap syndrome” this is best viewed as a distinct entity characterized by the presence of antibodies to a particular ribonuclear protein (RNP).

ILD---incidence and prevalence poorly defined, due in part to confusion about the definition of the disease itself.

PAH---similar to that of SSc.

Pleural disease---in contrast to SSc, (and more like SLE), effusions are common.

Diffuse alveolar hemorrhage (DAH)---similar to SLE.

Rheumatoid arthritis (RA)

ILD---multiple histologic patterns.

Pleural disease.

Airway disease---cricoarytenoid arthritis, bronchiolitis obliterans and multiple other processes in the lower airways difficult for clinicians to distinguish from COPD and often coexisting with ILD.

Nodules---solitary or multiple.

Caplan syndrome---RA + any of a variety of pneumoconioses, manifesting as the rapid appearance of multiple pulmonary nodules, some with cavitation. Mainly of historical interest.

Systemic lupus erythematosis (SLE)

Pleural disease.

Acute lupus pneumonitis---rare; high mortality; may present like pneumonia or as ARDS. Treated with steroids, associated with anti-double stranded DNA.

Chronic ILD---its strong association with anti-SSA has led to dispute over whether these patients represent Sjogren's or overlap.

Shrinking lung syndrome---sometimes confused with, but to be distinguished from, vanishing lung syndrome. Physiologically characterized by low lung volumes and elevated diaphragms. Unclear whether myopathic, neuropathic or both.

ARDS---not a specific entity, but can occur from acute lupus pneumonitis, DAH or infections. (See here regarding “atypical” forms of ARDS, ie ARDS as a manifestation of specific diseases).

Pulmonary vascular disease---PAH if overlap syndrome, PTE if anti-phospholipid syndrome coexists.

Sjogren syndrome

Airway disease---dryness, lymphocytic exocrinitis.


PAH, pleural disease rare.


Pneumonia risk, said to be “aspiration” pneumonia. “Aspiration” is in quotes because of the meaningless way in which the term is commonly used in pulmonary medicine. Note that almost all bacterial pneumonias, even in healthy hosts, are caused by aspiration of colonizing flora. It would be more accurate to say that patients with myositis are at risk for pneumonia due to weakness of pharyngeal muscles and difficulty clearing secretions and colonizing flora. Note too that some of these pneumonias are due to treatment related opportunistic infections. In those cases inhalation or hematogenous seeding are more likely to be the inciting event as opposed to aspiration.

ILD---I blogged about myositis associated ILD here.


The main ones are Wegener's, microscopic polyangiitis (MPA) and Churg-Strauss vasculitis.

Anti-phospholipid syndrome


ARDS (when anti-phospholipid syndrome becomes catastrophic anti-phospholipid syndrome).

Ankylosing spondylitis

Restrictive defect---fusion of chest wall joints.

Apical fibrocystic disease---superinfection of apical disease is common, particularly with non-tuberculous mycobacteria and aspergillus.

Monday, October 24, 2011

Health risks in adult survivors of childhood cancer

Although this NEJM review focuses on survivors of ALL some general principles apply. This is of peripheral interest to hospitalists. The review is free full text, at least for now.

Points of interest:

The care of these patients requires special expertise. Pediatric oncologists provide this service and there are “survivor clinics.” (See this link).

In addition to an increased risk of solid tumors survivors of ALL, though at virtually no risk of recurrence of ALL, are at increased risk for late post treatment myelodysplasia and AML. The degree of this risk depends on the original agents used.

Patients have an increased risk of obesity and metabolic syndrome. This may relate to dysfunction in the GH axis in those who received cranial irradiation.

Those given anthracyclines have an increased risk of heart failure that increases over time, even when lower doses were used.

More about the article from the NEJM blog.

Caffeine and cardiac arrhythmias

Is it bad for you? Probably not in moderate amounts according to this review.

Is baclofen ready for prime time in the treatment of alcohol withdrawal?

This has been a subject of numerous recent papers. A study just published in the Journal of Hospital Medicine looked at whether baclofen usage could reduce the required dose of lorazepam given in a symptom driven protocol. Here were the findings:

The 44 subjects who developed symptoms of AWS were randomized to baclofen or placebo. Thirty-one subjects (18 baclofen, 13 placebo) completed 72 hours of assessments, either entirely as inpatients or with outpatient follow-up. The need for high doses of benzodiazepines (20 mg or more of lorazepam over 72 hours) to control AWS was less likely in the baclofen treatment group (1 of 18) than in the placebo-treated group (7 of 13) (P = 0.004).

We need to note a couple of things about this study. First, it was very small. Second is the question of outcomes. The outcome of this study was the dose of lorazepam required in a symptom driven protocol. But is the requirement of more lorazepam necessarily a worse outcome? What we'd really like to know is whether patients receiving baclofen remain more alert, are better able to eat, are less likely to require mechanical ventilation, have shorter ICU and hospital stays and, of course, have a lower mortality. The baclofen story will unfold. This was of interest not a study to change practice.

This raises the whole subject of adjunctive drugs for alcohol withdrawal such as baclofen, clonidine and antipsychotics. I recently searched this question in UptoDate. The authors there favor sticking with benzos, and giving enough, and generally avoiding adjunctive medications.

HT to Hospital Medicine Quick Hits.

Saturday, October 22, 2011

Judge guidelines by the science, not the company they keep

Tom Sullivan, blogging over at Policy and Medicine, recently noticed another flurry of activity in the long running debate about conflicts of interest. He cited an article in BMJ and a post at Forbes blog. His post is titled Coordinated Ad Hominem Attacks on Physician-Industry Relationships in Guideline Development: The Next Frontier? It opens with:

Recently, we saw concerted attacks on clinical guidelines committees, but interestingly, not on the science coming out of them. Instead, the attacks were focused on whether the writers of those guidelines have a financial interest in the area they are working in.

That, I believe, is the crux of the whole debate. There has been a lot of criticism of practice guidelines based entirely on the financial interests of the guideline writers. I agree with Sullivan that it really is an ad hominem attack (and therefore fallacious) because virtually all of the criticisms of guidelines that have been written are based not on the science or the primary sources referenced in the guidelines but on the company they keep. I have read many such attacks. Not a single one that I can recall referenced the actual science. The recent examples cited by Sullivan are true to form.

As responsible clinicians we have an obligation to be scholars. That means being familiar with scientific underpinnings derived from primary sources. It's the only frame of reference we should be using. The arguments of the guideline attackers, devoid of any scientific appeal, always come up empty. More than that, they fail to take into account patient outcomes. Study after study has shown improvement in important patient outcomes, including mortality, associated with adherence to a variety of guidelines. Of scores of guidelines in existence only two have been associated with negative outcomes. (In the case of one of these the flaw is well understood, and known not to be related to industry influence).

I'm not suggesting we follow guidelines uncritically. That would be a violation of the first principles of evidence based medicine, which appeal to the primary evidence and ask the clinician to apply this evidence to individual patient attributes. But there's no principle of evidence based medicine that tells us to first drop down to the bottom of a scientific paper or guideline document, look at the author's disclosures then either believe it or reject it based simply on that.

You can't trust a guideline implicitly but the remedy for that problem is to go to the primary sources and look at the actual science. The Internet is easy to use and PubMed is free. To judge a clinical guideline by the simple litmus test of financial interest is not only an ad hominem fallacy, it is intellectually lazy.  We need to do better.

Friday, October 21, 2011

Corticosteroids in septic shock---current state of the controversy

The use of steroids in patients with septic shock is controversial. The Annane study in 2002 showed benefit and was followed by enthusiasm for the use of steroids. But several years later the CORTICUS paper , the largest study to date, showed no benefit of steroids. That didn't settle things, however. The CORTICUS patients were much less ill than the Annane study patients and debate continued. A meta-analysis published a short time later suggested benefit from steroids. Some patients clearly benefit. The trick is in knowing which ones. There is no basis in high level evidence to come down strongly for or against steroids as a general recommendation. The Surviving Sepsis guidelines recommend consideration of steroid use in patients poorly responsive to pressors after volume optimization.

Where do things stand now? In a recent commentary (HT to the EM Crit blog) in the journal Critical Care Dr. Paul Marik attempts to synthesize the evidence and offer updated recommendations. His recommendations are similar to those of the guidelines. As a specific indication of poor responsiveness to pressors, he considers the requirement of 0.1 mcg/kg/min of norepi (or the equivalent) as a sign that corticosteroids are indicated.

What is thrombotic storm?

Thrombotic storm is a syndrome of rapidly recurring venous or arterial thromboembolic events first described by Dr. Craig Kitchens at the University of Florida, in this article:

Six cases are presented demonstrating key features of what may be termed thrombotic storm: (1) an underlying hypercoagulable disorder; (2) a provocation to initiate thrombosis; (3) rapid development of new thromboses; (4) response to prompt use of thrombolytic agent or anticoagulant therapy; and (5) remarkable good long-term prognosis if the cycle of thrombosis is interrupted. Continued activation of coagulation by fresh thrombosis is hypothesized as the cause of the syndrome, which may explain its control by anticoagulants. Whereas these unusual patients’ courses most likely represent only an extreme of hypercoagulability and not a new disorder, their characteristic behavior warrants attention.

The underlying hypercoagulable disorder was not always identified although strongly suspected to be present based on underlying clinical behavior.

Recently Kitchens and his group have published an update in the same journal. In their series to date the presence of antiphospholipid antibodies was the most common underlying hypercoagulable state. Those patients resembled catastrophic antiphospholipid syndrome. There appears to be some overlap between catastrophic antiphospholipid syndrome and thrombotic storm. Catastrophic antiphospholipid syndrome may represent a subset of thrombotic storm.

The word “storm” is apt. All storms eventually pass. Thrombotic storm is remarkable in that it remits with the institution of effective anticoagulant treatment.

Thursday, October 20, 2011

Activated protein C for severe sepsis: what's the current status?

Activated protein C (Xigris) is given a soft recommendation for use in patients with severe sepsis at high risk of death in the latest version of the Surviving Sepsis guidelines.

In recent years it has fallen out of use due to various controversies, including allegations of undue Pharma influence, concern about bleeding risk and mixed evidence since the publication of PROWESS concerning optimal patient selection.

So what is its current status?

Two recent articles caught my attention, which I haven't blogged before. This large propensity matched retrospective cohort study showed a mortality benefit almost identical to that observed in PROWESS. This updated Cochrane review, on the other hand, found no evidence of benefit.

So for now the drug remains controversial. We'll have to wait for the results of Prowess-SHOCK, in which it is studied in the sickest of patients, to get better answers.

Meanwhile in a recent podcast at the EM Crit blog sepsis expert Emanuel Rivers, the originator of early goal directed therapy (EGDT) was asked about the controversy. He cited the increasing evidence that use of Xigris is time dependent, and blames some of the negative results on giving the drug inappropriately late, remarking that inappropriately late use in the course of sepsis has unfortunately “done this drug in.”

Upper extremity DVT

Here is a recent review in the American Journal of Medicine.

Some key points:

Deep veins are defined as brachial, acillary, brachiocephalic, IJ and subclavian veins. Basilic, cephalic and EJ are considered superficial.

Secondary UE DVT is the most common presentation, and is usually due to central IV catheters including peripherally inserted (PICCs). Other risk factors such as malignancy are included.

Primary UE DVT is less common and also known as Paget-Schroetter syndrome. It may be precipitated by vigorous arm activity and may be associated with thoracic outlet abnormalities.

The diagnostic approach is similar to that for LE DVT although supporting research data are relatively sparse.

There appears to be a lower incidence of concomitant PE although sequellae are otherwise similar to those of LE DVT.

Treatment, including duration, is similar to that for LE DVT.

Because cosmetic and comfort issues are more likely to surface with UE DVT than for LE DVT, thrombolysis (systemic or regional) or catheter extraction are often considered. However, data are limited and based on the evidence, at present, such techniques are recommended mainly for patients with massive clot burdens.

Could your patient's seizures be due to hyponatremia?

The answer may be more difficult than you think according to this post from Renal Fellow Network.

Wednesday, October 19, 2011

Electronic medical records disappoint

---according to this essay :

As with any system, an EHR produces the outcome for which it was designed. The most widely used EHRs are designed for auditing, compliance and billing. They are not designed for more efficient patient care in the office, better communication among providers, or the collection of group data for quality improvement efforts. As yet, there is no evidence that the use of an EHR provides less expensive care and/or better outcomes.

Indeed there is no evidence, only dogma, and it's dogma that's being forced down doctors' throats. Go to the link and read the rest. Not just anyone could speak such heterodoxy but this article was written by none other than the president of the American College of Physicians.

HT to Health Care Renewal.

Preoperative anemia associated with worse surgical outcomes

---in this much talked about study. Although the differences in outcomes between those with and without anemia held up to severity adjustment it's tempting to think that anemia was a marker of something else going on in the patient (infection, inflammation?) that may have impacted outcomes rather than a reduced oxygen carrying capacity. This will be studied further. We don't have an answer about whether a more aggressive transfusion strategy is warranted.

Via Hospital Medicine Quick Hits.

Subglottic secretion drainage

---decreased VAP in this meta-analysis.

Necrotizing soft tissue infections

Here is an excellent review of the topic.

A few key points---

Definitions and classifications:

The FDA classification is broken down into complicated and uncomplicated skin and soft tissue infections. Complicated soft tissue infection basically means any infection at deep enough levels to require surgical intervention. However, the FDA designations exclude necrotizing infections.

Necrotizing infections are designated types I, II, III and miscellaneous.

Type I infections are mixed infections, with both aerobic and anerobic organisms. Their pathogenicity is driven largely by host factors especially diabetes. Fournier's gangrene represents but one example and is defined by location in the perineal or genital area.

Type II infections are monomicrobial, involving certain necrotizing (popularly known as “flesh eating”) strains of either beta strep or Ca-MRSA. Host factors play a lesser role.

Type III infections are rare, very severe infections, with gas gangrene as the prototype.

Miscelaneous infections include water borne infections with Vibrio vulnificus as a prototype.

Ancillary diagnostic tools, including imaging and clinical scoring, as well as a table of antibiotic recommendations, are included in the article.

Hip fracture comanagement

Here are some tips presented at a session at the SHM 2011 national meeting.

Pradaxa, bleeding and the kidney

Here's a recent review and related post at The Kidney Doctor. Background here.

Nephrotic syndrome

Nephrotic syndrome is a podocyte disorder and is discussed in this review.

Points of interest---

The podocyte provides structural support for the glomerular capillary tuft and is important in the secretion of the collagen network of the GBM:

Failure to secrete this network results in a range of hereditary nephropathies, the type IV collagenopathies.
Type IV collagenopathies include Alport syndrome, nail patella syndrome, and thin basement membrane disease; all can be considered podocyte disorders.

The pathophysiology of podocyte injury explains why patients do not generally have active urinary sediment, in contrast to the glomerulonephritides:

Subepithelial immune complexes (as in MN) or podocyte injury usually do not lead to leukocyte recruitment and inflammation, but rather disrupt the GFB.
Typically, urine sediment is devoid of leukocytes and erythrocytes.
Disruption of GFB leads to proteinuria.
In contrast, injury to mesangial or endothelial cells, which are in direct contact with blood (containing leukocytes, complement, and inflammatory proteins), typically leads to inflammatory kidney disease (nephritis) with active urine sediment.

All patients with nephrotic syndrome are at risk for renal vein thrombosis, but especially those with membranous nephropathy. RVT may be asymptomatic, present with pain, or as a sudden decline in renal function.

Minimal change disease (MCD)

Inflammatory mediators---> podocyte injury---> foot process effacement.

10%-15% of nephrotic syndrome (NS) in adults, 90% in kids.

Secondary causes of MCD: drugs, neoplasia, others.

Focal segmental glomerulosclerosis (FSGS)

Pathogenesis complex and varied, but involves podocyte injury and podocytopenia.

Most common cause of NS in adults overall.

Can progress to ESRD.

Can be idiopathic but many secondary causes.

Collapsing glomerulopathy

Controversial whether a type of FSGS or a distinct entity.

Disease association: HIV.

Membranous nephropathy (MN)

Subepithelial immune complex deposition.

Primary or secondary. Many secondary causes.

Primary: antibodies enter subepithelial space and bind to podocyte antigen.

Secondary: a) circulating immune complex dissociates then reforms in subepithelial space or b) antigen lodged in subepithelial space and subsequent immune response. Specific antigen (tumor antigen, infectious particle, etc.) depends on underlying disease.

Long list of secondary causes (drugs, infections, neoplasia, rheumatic diseases).

Most common cause of NS in elderly, caucasians.

Urine: no “active sediment” but microhematuria common.

Miscellaneous disorders

Collagenopathies, etc., see above.

Via Nephron Power.

Tuesday, October 18, 2011

Diagnostic skepticism---a victim of competing interests

When I first read this post by Dr. Robert Centor at DB's Medical Rants about diagnostic skepticism I wondered why anyone would need to write about such a no-brainer. After all, rigorous skepticism is ingrained in us as physicians, right? But as I thought about it I realized that skepticism is lacking from day to day practice. As a hospitalist (and I'm sure this is true for other physicians) I have observed many things that conspire against diagnostic skepticism. These include the pressures of time, the performance movement and other competing interests as well as even the electronic medical record.

Some of the performance measures, particularly those that are time dependent, may cause the clinician to rush to a diagnosis. How many patients in the ER, for example, have been inappropriately diagnosed as pneumonia because the physician was in a rush to satisfy the time-to-antibiotic performance metric? How many patients with chest pain and ST elevation have been inappropriately rushed to the cath lab with pericarditis or even benign early repolarization because of the pressure of door-to-balloon time? Diagnostic skepticism does take time. Time to think.

Time pressures may also be driven by a variety of other administrative agendas. With all the emphasis today on rapid ER throughput, for example, patients need a quick diagnostic label as their ticket to get them admitted in a timely manner. These labels are administratively important but may lead to incorrect diagnoses. Other competing interests are about money, such as “clinical documentation” and DRG coding.

The EMR itself may enable inappropriate diagnosis and undermine diagnostic skepticism in a number of ways. Electronically generated problem lists and note templates are tied to the decades-obsolete ICD-9 codes. You can always free text the patient's real diagnosis but it's more often the electronically generated garbage that will be forwarded from note to note over time thus perpetuating inappropriate diagnosis. So in order to employ diagnostic skepticism you have to sit down and think. Not only that but you may have to undo some of the electronic garbage generated by the EMR. It all takes time. Time that hospitalists all too often don't have.

It's all made worse by the conflicting incentives. As a hospitalist, for example, what are some of the metrics you are held to? ER throughput, discharge time, patient satisfaction, coding and length of stay to name a few. These all take time away from diagnostic skepticism. Hospitalists are not incentivized for their diagnostic skills.

Our hospitalist leaders (except of course for Dr. Centor) are not helping matters. Take a look at the Society of Hospital Medicine website. There are some great resources there but almost all of them are about system and administrative issues in hospital medicine. There is no focus on diagnostic skills.

So the notion of diagnostic skepticism is very simple. All you have to do is stop and think. Take a moment to consider whether the diagnostic label perpetuated by the EMR is really true. Who knows, you may experience a Dr. House moment. You may even save a life.

Monday, October 17, 2011

Obamacare's CLASS program on hold

What does it mean for the rest of the law?

Does the Joint Commission improve the quality of hospitals?

Well, a study just published in the Journal of Hospital Medicine implies that maybe it does. Investigators looked at publicly reported performance data for accredited and non-accredited hospitals. Here are the findings:

Hospitals accredited by The Joint Commission tended to have better baseline performance in 2004 than non-accredited hospitals. Accredited hospitals had larger gains over time, and were significantly more likely to have high performance in 2008 on 13 out of 16 standardized clinical performance measures and all summary scores.

While Joint Commission-accredited hospitals already outperformed non-accredited hospitals on publicly reported quality measures in the early days of public reporting, these differences became significantly more pronounced over 5 years of observation. Future research should examine whether accreditation actually promotes improved performance or is a marker for other hospital characteristics associated with such performance.

So does this mean that the Joint Commission improves the quality in our hospitals? No. Does it even mean that Joint Commission accreditation is associated with better quality? No. We need to examine some questions.

First, what about these non-accredited hospitals? Well I was surprised to learn from the paper that as many as 18% of hospitals are not accredited. The study compared those hospitals that were never accredited for the entire study period with those that were accredited the whole time. Although the authors were not explicit about it, the strong sense I got from reading the paper is that these were all hospitals which chose not to seek accreditation rather than those which tried and flunked year after year. It seems to me that accreditation by the Joint is a lot like a hospital's report on performance measures. They share almost identical incentives. Both amount to public report cards. So if your hospital administrators are not interested in one they're not likely to be as interested in the other, compared to hospitals that seek accreditation. In my view the study demonstrated nothing more than that.

But the fundamental flaw in this paper, to me, is that the authors seem to confuse performance with quality. I've blogged extensively before on this fallacy. The authors are smart folks (Bob Wachter among them). Surely they are aware of the difference. I only wish that in their discussion they had made a clear distinction.

HT to Hospital Medicine Quick Hits.

Friday, October 14, 2011

A rapid test for tetanus immunity status

The Tetanus Quick Stick.

ACE inhibitor induced small bowel angioedema presenting as an acute abdomen---CT characteristics

From the American Journal of Roentgenology:

ACEI-induced small-bowel angioedema should be included in the differential diagnosis when patients receiving ACEI therapy present with abdominal complaints and the following combination of findings on CT examination: ascites, small-bowel wall thickening, dilatation without obstruction, and straightening.

Symptoms resolve within a few days of stopping the ACEI.

Via Hospital Medicine Quick Hits.

Thursday, October 06, 2011

Approach to the patient with severely elevated blood pressure

So you're called by the ER doc, saying “I've got this previously healthy patient” (maybe he was seen for a minor injury) “and he had a blood pressure of 190/120, asymptomatic. I've given him a couple of doses of clonidine. He's been here for over four hours and I can't get it below 180/100, so I think he needs to come in to be observed and treated overnight.” We've all been there. That seems to be the knee jerk approach. But is it the best approach? How can the best evidence guide us?

It's all nicely outlined in a review published in American Family Physician, linked from a post at The Kidney Doctor. The review is concise, practical and evidence based, and one of those few full text articles from AFP freely available on line (BTW, AFP now has the most restrictive access policy of any journal I know. You normally can't even get the abstract unless you are a subscriber).

According to the review the first thing you need to do is to classify the patient's elevated blood pressure. That can be confusing. Although JNC 7 (JNC 8 is supposed to come out next year) gives the traditional hypertension classification of prehypertension and stages 1 and 2 hypertension it does not define the terms for patients who present with severely elevated blood pressure. To make matters worse the terms out there right now are widely misused. Terms like “hypertensive crisis” are applied to so many different clinical circumstances they have become meaningless. “Malignant hypertension” used to mean something specific but has suffered the same fate in recent years. Even worse is the clinically inappropriate terminology in the ICD 9 codes we're forced to use (those are, what, 40 years out of date now?).

The terms are defined and the classification is succinctly presented in the review. Severely elevated blood pressure of any classification is defined as being greater than 180/100. If it is entirely uncomplicated without historical risk factors (heart disease, renal disease) or acute organ damage it's called severe asymptomatic hypertension or severe uncontrolled hypertension. If it is associated with historic cardiovascular risk factors or renal disease it is called hypertensive urgency. Finally, the term hypertensive emergency is reserved for individuals with evidence of acute organ damage. (Now when the latter are present they may trump the absolute level of blood pressure in defining a hypertensive emergency, e.g. aortic dissection, acute decompensated heart failure).

So that's step 1. Classify the patient using current terminology. How often is that done?

The article goes a little into pathophysiology. For our purposes I'll just mention the most important core concept: autoregulation. A graphic of the autoregulation curve is provided in the article. It's a familiar concept to most of you. If not, get it emblazoned in your hippocampus. Raise BP above the autoregulatory zone and you've got hypertensive encephalopathy. Go below, and perfusion of vital organs is proportional to blood pressure and may drop below critical levels. Acute “normalization” of blood pressure in the ER may take the patient into that zone.

As far as treatment is concerned go to the article, but here is the Cliff Notes version. In general the recommendations for starting patients on meds or admitting them to the hospital are a lot more conservative than what is commonly done. If it is really a HT emergency they go to the ICU. In HT emergency the BP target, the specific agent and route are specific to the underlying condition and are beyond the scope of this post as they are for the article.

For HT urgency or severe asymptomatic HT there appears to be no specific evidence that you should treat them in the ER or send them out with a prescription. Of course then the approach may be dictated by the patient's social situation. If non compliance, poor social connections or lack of a PCP complicate the situation it may be wise to have the patient spend the night in the hospital. But if the patient is better situated, according to the best evidence s/he could be sent home without meds with plans for very early PCP follow up (24-48 hrs if HT urgency, a week if severe asymptomatic HT) at which time the PCP would further evaluate the patient and initiate oral medication.

Wednesday, October 05, 2011

Fake doctor notes in Wisconsin: civil disobedience or something else?

The other day physician and journalist Ford Vox wrote a piece on the ethics and consequences of the fake medical excuses written by doctors in the Department of Family Medicine at the University of Wisconsin. In saying that the docs failed to understand the impact of their action on public perception of the profession, he was right. In crediting them with an act of civil disobedience he missed the boat:

Members of the University of Wisconsin's Department of Family Medicine, including both residents and attendings, felt compassion for the educators' cause, and wanted to participate in the display, but rather than symbolically protest in the streets alongside the teachers, they decided to perform an act of civil disobedience: The doctors wrote out fake sick notes for the teachers, many of whom were falsely using illness as an excuse to attend the protests..

Wrong. Civil disobedience is open. Say you wanted to protest U.S. military spending. An act of civil disobedience might be to openly refuse to pay income tax. It would not be an act of civil disobedience to cheat on your income tax even if you rationalized that the morality of your use of the money would exceed that of Uncle Sam. If that's too rude an analogy it illustrates the point. The doctors in Wisconsin may have thought they were demonstrating for social change but they were being, to use not so rude a term, disingenuous.

Also consider that those who commit true civil disobedience are open in facing the consequences of their actions. The Wisconsin doctors were not. When approached by reporters and videographers the docs offered lame (and unintentionally hilarious) defenses or told the reporters to get lost. One wonders if any of them, especially the residents involved, counted the cost in advance.

This little stunt may have been the use of the profession as a vehicle for social change but it wasn't true civil disobedience.

Fake notes by real docs in Wisconsin---anything new?

It's been over six months since faculty members of the University of Wisconsin department of Family Medicine made a stir when they decided social activism should trump evidence based medicine in the treatment of individual patients, and made a public display of it. Things have been quiet in the blogs after an initial splash, so I decided to search for updates and found a couple of recent news items.

The status of things, it would appear, is up in the air as investigations are ongoing at two levels---the state medical board and the University. Details will eventually be made public.

Resistant hypertension

There's a nice post over at The Kidney Doctor on this topic along with several references. The AHA scientific statement on resistant HT can be accessed here.

Monday, October 03, 2011

Wernicke's encephalopathy

A post about Wernicke's at Renal Fellow Network prompted this literature search.

This article covers CT and MRI findings:

CT and above all MRI of the brain play a fundamental role in diagnosing the condition and ruling out other diseases. MRI is the most sensitive technique and is required in all patients with a clinical suspicion of Wernicke’s encephalopathy. Medial thalami, mamillary bodies, tegmentum, periaqueductal region, and tectal plate are typical sites of abnormal MRI signal. The dorsal medulla, red nuclei, cranial nerve nuclei, cerebellum, corpus callosum, frontal and parietal cerebral cortex are less common sites of involvement although they are more frequently affected in nonalcoholic patients. Paramagnetic contrast material may help to identify lesions not otherwise visible.

This review of WE focuses on imaging findings but also comments on clinical criteria and pathophysiology:

In view of the poor diagnostic performance of the classical triad, new classification criteria have been proposed. These criteria require two of four items including dietary deficiencies, oculomotor abnormalities, cerebellar dysfunction, and an altered mental state or mild memory impairment [13]...
Alcoholism does not directly cause thiamine deficiency, although it may induce such deficiency because of its frequent association with malnourishment. More specifically, the low thiamine absorption rate at the mucosal level, the impaired hepatic function, and the alcohol-related raised thiamine metabolism together may lead to the development of chronic thiamine deficiency [14]. Thiamine-deficient membranes are unable to maintain osmotic gradients, which results in the swelling of intra- and extracellular spaces.

From the discussions in this and other reviews it is apparent that the mechanisms by which thiamine deficiency produce the brain lesions of WE are multifactorial and poorly understood.

This fascinating review explores the controversial and sometimes confusing overlaps between WE, Korsakoff Syndrome (KS) and non-thiamine related alcohol dementia.

This review highlights the increasing recognition of WE as a complication of bariatric surgery.

This article reviews several aspects of the clinical findings, natural history and pathology of WE and KS. It emphasizes the under-diagnosis of WE. This under-diagnosis is due to the fact that only 10-15% of patients present with the classic triad (eye findings, ataxia and mental status changes).

This case series illustrates that WE may be associated with a characteristic acid-base disturbance, combined respiratory alkalosis and wide anion gap metabolic acidosis such as is seen in salicylate intoxication.

Here is a review of the emergency prevention and management of WE in acutely ill patients emphasizing common pitfalls. The discussion of early recognition deals with non-WE presentations of thiamine deficiency such as cardiovascular collapse, heart failure, lactic acidosis and acute abdominal symptoms. Also listed are non-alcohol related conditions now known to be associated with WE: hyperemesis gravidarum, TPN without adequate thiamine supplementation and post-bariatric surgery states. The pathophysiology and biochemistry of WE are described as complex, and the brain lactate accumulation mentioned in the post at Renal Fellow Network is probably but one of several mechanisms of injury.

A 2007 review from Lancet Neurology cites an even wider variety of non-alcohol scenarios that can be associated with WE including fad diets and nutritional imbalances of various sorts, magnesium deficiency (it may induce refractoriness to administered thiamine) and various situations in cancer patients. The review makes the point which, though obvious, bears mention, that WE cannot reliably distinguished from drunkenness. The optimal route for thiamine administration is IV. Although deemed controversial in some of the other reviews this one makes more specific statements about dosing. 100 mg IV daily is common practice, and while this may be appropriate for prophylaxis in at risk patients, the review recommends 500 mg IV tid for patients with strongly suspected or established WE to prevent KS, other brain damage or death. (A banana bag may not be enough!).

The profile of WE has changed. During my training, as now, the diagnosis of WE was primarily clinical. Confirmatory diagnostic testing, however has changed. Then it was done by blood testing (thiamine levels or red cell transketolase assays) whereas now MRI is the diagnostic modality of choice. Neither is useful in the acute situation, as thiamine repletion must begin in the emergency setting before confirmatory test results are available. The epidemiology has changed. Because parenteral thiamine administration in the ER has become almost knee-jerk in alcoholic patients, we may be seeing less alcohol-related WE in comparison to other circumstances such as hyperemesis gravidarum, post-bariatric surgery states and other instances of disturbed nutrition.