Monday, December 2, 2013

Women, ACS, and "atypical" symptoms - new study

I was excited to read the new study "Sex-Specific Chest Pain Characteristics in the Early Diagnosis of Acute Myocardial Infarction." My excitement was tempered when I saw Ryan Radecki's take on the article, however.  

"Fundamentally flawed" - ouch.

True, the study by Gimenez et al. cannot prove, because of study design, that men and women have similar rates of atypical symptoms of ACS. However, I’m not sure that this makes the study fundamentally flawed, let alone “comical.” It’s too easy to criticize a study for not being a highly-powered, well-controlled (yet externally valid!) “gold-standard” investigation. But how do we proceed until that rare trial is conducted? In the meantime, why not analyze this study on its own terms? 

Most importantly, this was a prospective study of undifferentiated chest symptoms. Too much of the literature that is cited on this topic uses registry data, or only enrolls patients with confirmed ACS, or interviews subjects long after the onset of symptoms. While the study by Gimenez et al. may not be perfect, its methods are far more robust than most other literature in this area. Importantly, the results triangulate well with other kinds of evidence out there.

1. In most diseases, women have the same symptoms as men. 

How many other diseases are said to show a gender-based difference in presentation? Not many, it seems - I’m pressed to come up with another example where the literature even suggests a significantly different symptom complex.

Indeed, many studies point to the absence of differences in various condition. For example, men and women with pulmonary embolism seem to have similar rates of various symptoms. Interesting, since PE often produces the same sort of vague, poorly localized symptoms as ACS. Along the same vein, at least one study suggests that pancreatitis presents much the same in women as in men

Well, instead of symptom differences in heart attacks, how about “brain attacks?” Again, it seems that there are few clinically significant differences in the way men and women describe their stroke symptoms.

Even in appendicitis, a condition that manifests in protean ways, and involves anatomy that sits right next to gender-defining organs in the pelvis, it seems that women have pretty much the same symptoms as men!

So, if there are few (no?) other conditions in which men and women are understood to have clinically significant discrepancies in their symptoms, why should we believe that heart disease should be the rare (sole?) exception?

2. Evidence from the cath lab suggest men and women have similar symptoms. 
Balloon occlusion during PCI is, essentially, a temporary MI. The lumen is occluded for however long it takes to open the artery and deploy the stent, and people can have significant ischemic symptoms during this period. This makes for a great study setting - although we aren’t studying symptoms associated with ACS sensu strictu, we are able to prospectively survey the patients about their symptoms not only after artery occlusion, but before and during as well.

A team in Japan did just that, and found that men and women had chest pain-free occlusions at about the same rate, 35%. Another study, in Canada, also found no difference between men and women in the cath lab.

3. Many studies in this area are not designed to address the crucial question. 
Most of the data looking at men’s and women’s symptoms in ACS come from studies that enrolled patients with confirmed ACS. For example, in the Gimenez et al study, they cite a number of studies that suggest that women present with atypical symptoms more often. Unfortunately, Goldberg 1998, Goldberg 2000, and Dey 2009 all examined only patients with diagnosed ACS. As such, they can answer the question “In patients with an existing diagnosis of ACS, do women and men have different symptoms?” They cannot speak to how to approach the patient with undifferentiated chest pain. 

For example, many of these retrospective studies find that women describe back pain somewhat more often than men. But, what if women without ACS also describe back pain more often than men? If this were the case, back pain would not really be useful as an “atypical” symptom. 

It's worth pointing out that one of the few prospective studies, looking specifically for atypical symptoms in men and women, found that women presented typically more often than men!

Wrapping up...
So, while this study may not be the last word on the subject, it provides yet another high-quality element of evidence pointing towards the same conclusion: Both men and women - old and young, diabetic and not - can have typical or atypical symptoms of ACS.

Sunday, July 21, 2013

Lime's Disease - some underappreciated elements.

This is not meant to be a comprehensive review of B. burgdorferi, the life cycle of I. scapularis, or all the other stuff that belong in a thorough introduction to Lime's disease. Instead, I just want to highlight a few aspects that seem to have been under-appreciated.

First, an exacting review of the anatomy of I. scapularis...

This depiction is actually more realistic than how some CT politicians view ticks and Lyme disease.

#1 The "bulls-eye" rash, while classic, is uncommon.
Anytime a sign is described as classic, you ought to expect that,
basically, you will never see it. My rule of thumb when being pimped about how often you see a "classic sign" is to reply "Well, recent research, ah, I believe, shows a lower rate in the modern era, ah, about 15% I believe."

And the literature on erythema migrans (EM) backs me up. A "bulls-eye" pattern, with central-clearing, may have been more common years ago, when Lyme took weeks to diagnose. In contrast, a study from 2002 found that only 9% of confirmed EM had central clearing. Instead, the majority either were homogenous, or were darker centrally!

In fact, central clearing occurred at the same rate as rashes which had vesicles or a blue center.

So, all these are erythema migrans:
Central vesicles (source)
Central raised punctum (source)
Triangular, rather than round (source)

What should you look for, if not a bull's eye? Five key elements are:

  • Blanchable erythema,
  • Flat, with non-raised border. 
  • Large, and rapidly expanding (20 cm2/day).

#2 Don't routinely get "Lyme tests."
The patient lives in suburban Connecticut, it's July, they describe "flu-like" aching and chills, and you find a 15 cm diameter homogenously erythematous rash on their back. You're done - 2 weeks doxy 100 BID, and go see the next patient!

But what about a test, "just to make sure?" Hey, we're always getting tests. We order BNPs on patients who are on BiPAP and getting 400 µg/minute of nitro, we get a troponin on the patient being rushed to the cath lab for anterior "tombstones," we get a white count in, well, everyone. So why not order a test for Lyme?

Because they don't work well. Some of the pitfalls are:
  • In early localized disease (i.e. EM) about 50% of patients will not yet have a rise in IgM levels.
  • About 5% of the population can  have a positive ELISA test at any given time.
  • In the absence of a supportive history or clinical signs, a positive IgG just indicates past exposure.
  • Elevated IgG, and even IgM, levels can be seen for a long time after successful treatment.
If the rash is equivocal-looking, you can get active and convalescent assays, but this is not routine, and should require discussion with the PMD.

And I enjoy trying to page PMDs almost as much as I enjoy this new yogurt flavor!

#3 Lyme carditis
An otherwise healthy 35 y.o. male comes to the ED with severe presyncope, after having been found to have a heart rate of 30 in the walk-in clinic. He admits to having been told by a coworker at his landscaping job that he had a big red rash on his back 3 months ago (in July), but he never saw it himself. The blood pressure is 80/40, and the ECG shows a complete heart block with a narrow QRS.

How bad is this? I mean, complete heart block - yikes. What should we do right now? Does he need a permanent pacemaker? Will the echo show a nasty cardiomyopathy? How bad is the mortality?

Ok, in order:
  1. Not that bad. These blocks usually last under a week, once antibiotics have been started.
  2. Not a question... But AV blocks are very common in Lyme carditis - about 50% of patients.
  3. Put on pacer pads, order some ceftriaxone, and don't let him walk to the bathroom!
  4. Unlikely. Permanent conduction defects are rare, even though some folks need temporary pacing.
  5. Maybe in the short-term there can be some "reduced squeeze," but the long-term prognosis is very good.
  6. Almost unheard of: a recent review only found two case reports that plausibly link a death due to Lyme disease
Lastly, atropine is not felt to do much, good or bad.

#4 Prophylactic Doxycyline
If a deer tick has been for at least 36 hours, and the patient can take antibiotics within 72 hours after tick removal, and we're in Connecticut (i.e. a Lyme endemic area), the the patient should get doxycyline 200mg PO once.

There are a few wrinkles in this, however. For example, you can't do this for kids - there is no data for prophylactic-dose amoxicillin. 

But most importantly, you have to know the risk-benefit numbers. First, what is the risk of developing erythema migrans after a tick bite, and how much does doxy help? The key NEJM study found:

It looks like most deer tick bites, even in Westchester, NY (a Lyme endemic area), do not result in EM. The risk tops out at about 10% for a somewhat engorged nymph, and plummet for the other categories. The one-time dose of doxy drops that rate down to a little over 1%. That's a pretty decent benefit.

Well, how about the risks of prophylactic dosing? 

A 6% risk of vomiting, and 7% abdminal pain? Hmm.

So, another way of looking at it, the patient potentially has a 90% chance of having nothing happen (if no prophylactic dose), versus a 6% chance of being sick as a dog from the doxy. That's the choice!

The Bottom Line
There's a big fear about Lyme disease in Connecticut, and plenty of people work hard every day to make sure that the paranoia doesn't die down.

Like one of our senators, who investigated and sued the Infectious Disease Society of America
(spending at lot of CT tax money in the process), and was ultimately 100% in the wrong. ***
So, as an emergency doctor in this wacky state, you should know this disease pretty well, so you can identify and treat "Lime's disease" appropriately. You can download the excellent IDSA guidelines for definitive information, or check out the CDC website for clinicians as well.

***  My own political views are not represented here, just a medical perspective. So, in order to balance out my criticisms of a Democrat, let me point out that no political party has a lock on pandering to the "chronic Lyme" folks. To highlight a recent example:

He planned to "improve synergy" in treating Lyme. Cool.

Friday, March 29, 2013

Symptoms in men & women during PCI ballon occlusion

The trouble with "chest pain" is that it's, well, pain. The experience is subjective, bound up in the context of prior experience, current emotional state, and comorbidities, as well as the actual nociceptive stimulus. 

Not Garfield.
Frankly, this complex topic does not get any simpler when we consider the question of whether men & women report significantly different symptoms during ACS

First off, I harbor some skepticism about whether such a significant difference exists at all. For all the talk of "men present typically, women present atypically," there isn't a great deal of evidence that women have substantially different symptoms with pancreatitis, pulmonary embolism, CVA, or even appendicitis. Why should the heart be such a radically different organ? 

Nonetheless, the latest big study on the subject suggested that the rates of "chest pain-free" MI are different between the genders: 31% of males versus 42% of females. However, this was a registry study, with all the usual limitations, and so questions remain. But what can you do? It's not like you can take a group of women, a group of men, give them both MIs, and record their symptoms...

Or can you?

"Gender Differences in Symptoms During 60-Second Balloon Occlusion of the Coronary Artery"
Well, sort of.

Japanese researchers decided to approach this issue prospectively. They enrolled 110 men and 80 women who were scheduled to have PCI for stenting of a single stenotic native coronary artery. None of these patients were having any active symptoms (let alone a STEMI) prior to undergoing PCI. They figured that since the balloon inflation required to deploy the stent causes, essentially, a transient total occlusion, it might prove to be a good model for demonstrating the different symptom expression between genders.

The men and the women were fairly well matched, with the exception that the women were, on average, older than the men. The rates of comorbidities, as well as the target vessels, were similar, however.


The duration of balloon inflation during the PCI was standardized at 60 seconds, and the patients were interviewed immediately after balloon deflation for:
"presence or absence of the following symptoms during the balloon inflation: chest pain, toothache, jaw pain, neck/throat pain, shoulder pain, upper or lower arm pain, epigastralgia, occipital pain, back pain, dyspnea, nausea, and vomiting."
The first interesting point is that not everyone had chest pain when the balloon was inflated. In fact, 38/110 men (35%) and 28/80 (35%) women had no symptoms at all.

So, the genders were evenly matched for "pain-free" coronary occlusion. What kind of symptoms did the remaining 65% of men and women have?

Surprisingly, essentially all the men and women reported having "chest pain." When asked about additional symptoms, however, women reported more of the "non-chest" symptoms than did the men, although none of the individual elements reached statistical significance.

To summarize: Men and women had the same rate of symptom-free coronary occlusion, and the same rate of chest pain. Women had more symptoms on top of that chest pain, however. The researchers concluded, nevertheless, that:
[N]on–chest pain symptoms during the 60-second balloon occlusion of the coronary artery were more common in women than in men, supporting the presence of the gender difference in myocardial ischemic symptoms.
I think we're making a mistake by focusing on a putative 11% difference in "typical" chest between genders, and should instead remind ourselves that 31% - 42% of patients with ACS do not have a typical presentation. In other words, the variation in symptoms between genders is dwarfed by the range of presentations within either gender.

Furthermore, the present study suggests that the difference in symptoms between men and women, when examined in a fairly well-controlled setting, may be trivial. These results are similar to those obtained by a Canadian group in 2011. Those researchers also employed a PCI setting to record patients' descriptions of symptoms during inflationOverall, they also found no significant difference between the genders.

Now, what do EM residents really want from the review of a new study? They want to know - Can you use the results of this study in the ED tomorrow? And the "trick answer" is that no special gender-based strategy is need or even helpful
 Just be concerned about ACS in everyone!