Just watched Amal Mattu (and Andy Neill’s) great EKG review, this time on (spoiler alert) — hyperkalemia! (If you’re not watching these videos, you’re missing out on free, amazing education from I think the best EKG teacher in the world.)
But I think there are two things worth mentioning:
You cannot — and should not — use an EKG to “rule out” hyperkalemia. I completely agree with Amal — if you see a bizarre looking EKG, you should think tox, potassium or calicum derangement (I like to throw LBBB in there too), but a normal EKG won’t rule out diddly squat in your patient. A few studies-in-point:
And secondofly — I’d have to disagree on Amal’s recommendation of bicarb.
Sodium bicarbonate does not work — or at least, does not work well, or on its own — for treating hyperkalemia. And in patients with fluid/volume issues (heart failure, renal failure — you know, the typical people who get hyperkalemic), I always worry about giving a big intravenous hyperosmotic sodium bolus to these patients (as my nephrology professor used to remind us — “water follows sodium”). (NB: This was brought to my attention by one of my co-chiefs, Kim Medlej, who finished a critical care fellowship last year at Harvard, and now practices in Lebanon, so all the credit is his.)
Quick summary: We’re all taught bicarb works within 30 minutes, by intracellular shift/exchange of potassium ions for hydrogen ions, yada yada yada. That really doesn’t appear to be the case. I think in the ED we’re sometimes taught to just give them an amp or two of sodium bicarb, but that appears to have NEVER been studied. In the crashing/dying patient, yes, I give sodium bicarbonate, but I’m otherwise skeptical of the benefit and worried about the harm.
All the studies have really looked at bicarb infusions over hours, and if there’s any change to be found, it’s maaaybe at the 6 hour mark (after 6 hours of bicarb infusion, in patients who are already getting dialysis). Other studies with bicarb infusions show no statistically significant change, either. (One study that took patients and put them on a high or low dose bicarb infusion for an hour actually found a higher potassium levels after the infusion.) Probably the best study
(Blumberg, 1992) found only a 0.5-0.7 drop, but they then attribute half the drop to the expansion of the ECF due to all the sodium the patients got.
Insulin definitely works. Albuterol works (but the studies are small and they usually give a good 10-20mg of it nebulized). There have been a few studies looking at combining bicarb + either of these other methods, and it looks like the bicarb probably DOES have some synergistic effect (it lowers the potassium more than just, say, albuterol alone). But by itself? Bicarb is probably pretty worthless.
Reviewing the literature, it seems like the insulin/D50, albuterol (? Lasix, not much literature on it) methods are the way to go. I know before I read this literature I felt better because I’d given the person kayexalate, or I’d given them bicarb, but really, the other methods are much more likely to keep the patient alive on the floor for 6 hours while they await their dialysis, without putting them into florid fluid overload.
To the stable, no dysrhythmias or severe symptoms patients I tend to give:
Regular Insulin 10 units IV with 1-2 amps D50
Albuterol 10mg nebulized
Calcium Gluconate 1-2g IV
Lasix if they make urine (pick your dose)
I’ve summarized the literature and we can send you the articles if you’re curious:
Burnell, 1956 http://www.ncbi.nlm.nih.gov/pubmed/13367188
Looks like this is where a lot of it started. Many articles from the
70s/80s cite this one. There’s very little on their methodology, but
they have some pretty cool graphs that show an inverse relationship
between pH and serum potassium concentration.
Schwarz, 1959 http://www.ncbi.nlm.nih.gov/pubmed/13629781
Case series of hyperK patients who had EKG changes who got better with
bicarb. (Some of them got calcium as well, others required “5-10 grams
of bicarb a day,” others got bicarb + blood transfusion.)
Fraley, 1977 http://www.ncbi.nlm.nih.gov/pubmed/24132
Methods: Took 14 hyperK patients, gave them bicarb infusions over 4-6
hours. Checked K every hour.
Results: Divided groups retrospectively into “constant pH” and
“changed pH” groups. Both groups showed decreases in their potassium,
~1.6-1.8mmol/L (never seen this significant of a drop reproduced).
Blumberg, 1998 http://www.ncbi.nlm.nih.gov/pubmed/3052050
Methods: Took 10 HD patients, checked their K (along with other labs),
gave them a bunch of different agents for changing K (bicarb, insulin,
epi drip, regular dialysate), and then checked their labs after an
hour. For bicarb, it was 8.4% in water, 4mmol/min, for 1 hour only.
They also tried a isotonic bicarb infusion of 1.4%.
Results: The K actually went UP after both bicarb infusions.
They conclude that bicarb didn’t work, but in the past it’s worked
over longer periods of time. So then they do …
Blumberg, 1992 http://www.ncbi.nlm.nih.gov/pubmed/1552710
Methods: Took 12 hyperK (>5.8) patients on dialysis, gave a bicarb
(8.4% in free water) infusion 4mmol/min x1 hour, then 1.4% bicarb in
water infusion 0.5mmol/min hours 2-6 and checked potassium levels
throughout the time on dialysis. Also checked an EKG.
Results: Average K was 6.0. K dropped at 4-6 hours, by 0.5-0.7, and
they believe that half of the drop is probably due to the huge sodium
load and increase in the extracellular fluid compartment.
Allon, 1996 http://www.ncbi.nlm.nih.gov/pubmed/8840939
Methods: Took 8 HD non-HyperK patients, put them through different
combinations to lower their K (bicarb infusion, saline infusion,
bicarb+insulin, saline+insulin, bicarb+albuterol, saline+albuterol).
Results: Bicarb or saline infusions didn’t work. Anything with insulin
or albuterol the combination worked, lowered them from 0.5-0.8,
depending on the group. Of note, bicacrb + albuterol worked better
than saline + albuterol (see Kim, 1997).
Kim, 1997 http://www.ncbi.nlm.nih.gov/pubmed/8852501
Methods: Took 9 HD hyperK patients, gave them separate or combined
bicarb infusions (1/2 hour long) along with nebulized albuterol,
checked K before and after. Thought maybe there would be
combined/synergistic effects of the two meds.
Results: Bicarb alone didn’t change the potassium. Salbutamol alone
dropped the K by 0.6, and salbutamol + bicarb dropped the K by 0.9.
Kaplan, 1997 http://www.ncbi.nlm.nih.gov/pubmed/9043534
Methods: Took 8 dogs, gave potassium infusion until they got
conduction disturbances, then backed down on the K, and gave either
bicarb infusion (1.05% over 1 hour), bicarb bolus (8.4% over 5
minutes, then saline), or “saline” therapy (hypertonic saline 8.4%
bolus + normal saline). Measured K before and after.
Results: Saline worked just as well as bolus. Infusion worked better
than both (but not statistically significant). Change was 1-2mmol/L.
Definitely doesn’t work short-term, but might still be
useful for temporizing hyperK. “It has now been clearly demonstrated
that short-term bicarbonate infusion does not reduce PK in patients
with dialysis-dependent kidney failure, implying that it does not
cause K shift into cells. Infusion of a hypertonic or an isotonic
bicarbonate solution for 60 mins has been shown to have no effect on
PK in dialysis patients, despite a substantial increase in serum
Rachoin, 2010: http://www.ncbi.nlm.nih.gov/pubmed/21661096
“When treating hyperkalemic patients, hospitalists
should use sodium bicarbonate to potentiate urinary elimination of
potassium and should consider administering it either with
acetazolamide or a loop diuretic, anticipating a lowering effect after
a few hours.26 It should be avoided in patients with volume overload
and anuria. Immediate translocation of potassium into cells is best
achieved by insulin and b-2 agonists.”
"I’d like a full set of vital signs, the patient put on the cardiac monitor with constant pulse oximetry, and 2 18G IVs in the antecubital fossae. Please ask the family and EMS to stay. I’d like the patient to get fully undressed and into a patient gown."
Fingerstick Glucose, FAST, Foley
Glucose, NG Tube
HCG, Human (Give pain meds, treat fever)
J=Tell Patient Plan
Inspired by the hacking of Amal Mattu et. al’s fantastic FOAM and Emergency Medicine pearl website, UMEM.org, I thought I’d quickly review some easy ways to backup your online content, from blogs to Twitter and Facebook.
Quick Guidelines to Backing Up
You can never have enough backups (you need at least two).
It’s best if you can automate some of the process, so you’ll keep your backups current.
The approach I’m going to suggest here either involves two backups: either:
Then save the archive to your hard drive, and either your external hard drive and/or your Dropbox account.
A quick word: please realize that with blogs, the template (which includes the style, design, images, etc) is separate from the content. If you don’t back up both, you’ll lose both. Finally, if you include images in your posts, those typically stored in an “image uploads” directory. Don’t forget those, too!
Also: blog content will typically be given to you as one big file (“Exported”), which contains all your posts and comments. You can then use this file to “Restore” your blog, or even transfer the content to another blogging service (“Importing your blog”).
Content: Follow this tutorial under “Export.”
Links: If you use the Links or Blogroll feature, follow this tutorial.
Template/Theme: You can’t really do this on Wordpress.com, but the site maintains a ton of backups.
Your Own Wordpress Blog
If you host your own Wordpress blog, there are a number of great plugins that will help automate the backup process for you. One that I use for theNNT is called Online Backup for Wordpress (there are a number of others that do the same thing). Namely, these plugins with automatically email you (or store online in the cloud) a full backup of your site for you at whatever frequency you want.
To install it, go to your Plugins → Add New menu, and search for “Online Backup”:
Then once installed, under Tools → Online Backup, you can set whatever options you want:
Run a manual backup
Set a scheduled backup (once a day, once a week, once a month)
Determine if you want to backup your content, your templates (themes), or both
Determine if you want to email the backup to yourself or store it in the cloud somewhere
How To Backup
The easiest way to backup is just copy a file from your own hard drive to an external hard drive. That’s it, done.
You can also do a sexier, incremental backup, but it takes some setup. (Incremental backup means the backup copy won’t re-copy every single file, which could take a long time, but will only update files that have changed; it will also store copies of the files at different times, so you can go back to a previous, saved version if you want.) I’m a Mac user, so I’m most familiar with Time Machine. In Windows 8, you can use the File History Backup option.
And then there’s Dropbox. Here’s why I think it’s the best tool:
Dropbox lives on your computer and in the cloud. Every file that you change, move, or add to your computer’s own Dropbox folder gets automatically uploaded to Dropbox’s servers as an online backup, and then also gets updated on every computer you’ve installed your Dropbox account to. Imagine you have a home computer and a work laptop. If you’re editing a file or updating a file at home, copy it (or just store it) in your Dropbox folder, and when you turn on your computer at work, Dropbox will download the most recent version to your work computer. No more emailing files to yourself!
Three other great features:
You can email a link of a file to anyone, and they can download it from your Dropbox account.
Dropbox also supports versioning or incremental backups. Let me show you. Here’s a file I’ve been working on. I recently updated it, but I realized I had a nice table in the original version that is long since deleted, and I can’t get it back. I can revert to the original document and get that table back:
You can have Dropbox sync folders outside of your computer’s Dropbox folder! This gets a bit technical, but it’s really useful. For example, I sync all my important documents — presentations, web design, my CV, manuscripts I’m working on — with Dropbox, but I don’t store everything in my Dropbox folder. I create “symlinks” to the Dropbox folder. Lifehacker can explain more.
There are many other great backup solutions similar to Dropbox out there, but I’ll admit, Dropbox is certainly my favorite.
Dr. Sasson arrived for her shift at 11 p.m. that Thursday. Filling in for another doctor at the last minute, she had skipped the two-hour nap she usually takes before a night shift. As her colleague ran through the list of patients, she thought, “I can just power through till 8 a.m.”
Amazing story, and amazing coordination from the hospital. A must-read for anyone in emergency medicine.
Mel Herbert said this before, and I’m going to say it a little bit differently: we’re sadly moving from the art of medicine to the click of medicine.
Let me explain.
I worry that we as a specialty (Emergency Medicine) are heading toward being as mind-numbing as the trauma services are nowadays:
You have trauma? You get pan scan.
You have chest pain? You get troponin.
Obviously there are other causes of chest pain besides acute coronary syndrome (more on that in another post), but once you’ve gotten to a point with the patient that you think you’re down to ACS vs. other atypical weird chest pain you can’t diagnose in the ED but is not life threatening (esophageal spasm, GERD, anxiety, etc), it’s so damn easy to just “Click: EKG, Troponin” (and we’re all reasonably worried about missing an MI for a whole host of reasons) that we just click, without considering pre-test probability in any of this.
So let’s compare these two articles:
You’ll notice that the Vancouver Rule got people to less than 2% risk, which is really, really good — and that with troponins, probably would have been even better.
The NEJM article also had “good” outcomes. But you’ll notice that these people got into the study if they were 40-74, and the Emergency Physician felt like they needed a workup for ACS. That’s the stinker, right there. Have we all just gone crazy (I’m included in the “we” here, too) and decided that everyone needs a workup for ACS? You take a low risk population (negative troponin, non-ischemic EKG, not a good story for ACS), and no matter what test you do on them, you’re not going to find much disease… because they’re low-risk to begin with.
You could take low-risk patients and do a coronary CT on them to decide they don’t have coronary artery disease, but you could also take low-risk patients and apply a silly test to them: you’re low risk if you’ve eaten cheese. (See, it works, because they’re already defined as low risk!)
But it’s just so damn easy to click and order a CT coronary angiogram on them that it’s going to happen to people who didn’t need the test in the first place.
I worry that we’ll start applying this test to low-risk patients — as the NEJM article suggests — and end up first, doing harm. Anyone else feel the same way?
“It is time for us to own up to our shortcomings in cancer screening, and we must start by acknowledging a hard fact: Doctors sometimes don’t know best. We are terrific at inventing new tests that can be performed on people. But we have been less good at figuring out which people should have them.”—The Trouble With ‘Doctor Knows Best’ from the NYT.
I made a mistake in my talk — which, I wish I could say was on purpose — but at least it quickly proves my points about the power of crowdsourcing and that you can’t lie on the internet.
6:57am: I post a tweet that my video is ready for viewing.
7:19am: Andy Neil corrects me — the first meme isn’t from Game of Thrones, it’s from Lord of the Rings.
3 hours later, and again 6 hours later, two other friends correct me on Facebook.
So I thought I’d give an overview, especially given the spirit of my talk.
The First Part
Due to a little tech snafu, the first few minutes of my talk didn’t get recorded. So instead, I did a simple Quicktime Screen Recording and re-presented it on my Macbook Air laptop.
Open Quicktime Player.
File → New Screen Recording
Make sure you turn on your microphone using the small ▼ arrow:
Click the Record Button.
Open your Keynote (or Powerpoint) presentation.
Start the presentation in Full Screen.
Talk through your presentation.
Close the presentation, click the Record Button again to stop recording.
You can now edit this clip directly in Quicktime, or export it to iMovie to edit there.
The Actual Talk
Stanford gives you three video feeds of your talk. I think the way HQMedEd has their talks laid out is by far the best, but I settled on something that focuses primarily on the slides, with a little overlay/Picture-in-Picture of me walking around, talking, to make it seem a little less boring. (Did it work?)
Here’s what you get from Stanford:
A 960x540 HD H.264 screen capture of the slides:
A 960x540 HD H.264 video of you with your slides:
And a 1024x400, HD H.264 side-by-side video of your talk:
You can make a Picture-in-Picture version using iMovie.
Frightening, depressing, but not all that surprising. JAMA article here. I can’t clearly tell from either article whether these women are:
on chronic narcotics for chronic pain and then get pregnant
in methadone programs and then get pregnant
abusing narcotics and pregnant
Because I honestly can’t see a whole bunch of OBs or non-OBs prescribing narcotics to pregnant women. I think I have once, and it was a pregnant woman with a nasty ankle fracture.
I certainly now see an epidemic — depending on the shift, honestly up to 10% of my patients are young people with chronic complaints on chronic narcotics. I’d have the say the most common (and most frustrating) is chronic abdominal pain. It’s usually but not always young, child-bearing-aged women, who — no surprise here — probably get pregnant at some point.
And, in the “These Two Reporters Should Have Filed Their Stories Together” category comes this piece from the NYT: E.R. Doctors Face Quandary on Painkillers, which is again no surprise to any of us in the Emergency Medicine world, but kind of nice to at least get some public validation of our frustration (especially that no one seems to be interested in helping us solve it):
“You can be faulted for not treating a patient’s pain — it’s considered the ‘fifth vital sign,’ ” said Dr. Abhi Mehrotra, the assistant director of the emergency medicine department at the University of North Carolina Hospitals. “We have to ask a patient’s pain, on a scale of 0 to 10, as well as document a reassessment of their pain after treatment.”
Dr. Benzoni, who is routinely rated on patient satisfaction and sometimes asked by management to explain a bad review, said that he feels at times as if he faces a no-win choice. “If you’re going to criticize me for not giving out narcotics, and you never praise me for correctly identifying a drug-seeker,” he said, “then I’m going to give out narcotics.”
Far be it from me to point fingers, but the whole JCAHO (sorry, “The Joint Commission”) “pain must be reassessed and is now a vital sign” bit probably got us in this mess. I’m all for treating people’s pain — easing suffering is part of why I went into medicine, and particularly Emergency Medicine — but it’s become a total quagmire.
And when I say “great,” I mean “depressing,” since it’s a bunch of big companies stifling inventors from getting better products into our hands: a retractable needle many years earlier, a lap towel that’s x-rayable.
I’m really pleased this is getting publicity, not because we can now blame mothers for something, but simply that people might start to accept that vaccines are probably not the cause of autism.
Readers of this little Tumblr space I’m sure are aware that this correlation being reported may or may not indicate causation, but anything that might make people start investigating any of the other changes in today’s society that could also be correlated with increasing autism (increased television, increased video games, increased high fructose corn syrup, increased travel, decreased exposure to an extended family, decreased time with both parents, increased divorce rates, increased Britney Spears fans) is a good thing in my book.
“They couldn’t imagine that there was anything worse than their dear mother and mother-in-law … aunt and grandmother dying. And yet, they found out — as so many of our patients eventually do — that there are worse things than having someone you love so much die. There’s having them die badly, suffering as they die.”—
I will definitely be framing things this way in the future.
There are things worse than your mother dying: dying badly. Suffering. Poked and prodded with needles every 6 hours in an ICU. Tubes in every orifice. Sedated so heavily she cannot see you, respond to you, or even know that you are with her at her side in her final days.