Written by Pendell Meyers, with edits by Steve Smith
Thanks to my attending Nic Thompson who superbly led this resuscitation
We received a call that a middle aged male in cardiac arrest was 5 minutes out. He was estimated to be in his 50s, with no known PMHx. He arrived with chest compressions ongoing, intubated, and being bagged. EMS report was that the patient had unknown down time with unwitnessed arrest, found initially in VFib arrest, defibrillated x1 followed by PEA arrest alternating with asystolic arrest during transport.
He arrived in PEA arrest with a slow and wide cardiac waveform during initial rhythm check, with faint cardiac activity on US but no palpable or visible femoral pulse on US of the femoral artery. Chest compressions were continued, and the patient was given 1 round of epinephrine, calcium, bicarb, glucose.
During the next rhythm check, we recognized the rhythm on the monitor as having features suggestive of severe hyperkalemia (sorry, monitor appearance not available). We then gave another 2g of calcium chloride.
Within 1-2 minutes ROSC occurred, chest compressions were stopped, a faint femoral pulse was palpated, and the femoral arterial line read 80/40. The following 12 lead was recorded.
|What do you think? What will you do?
There is a regular, wide complex, (mostly) monomorphic tachycardia. In a few leads you can see slightly different QRS morphologies, but overall I feel this can be called monomorphic. The QRS is VERY wide in some leads, and in V1 rhythm strip there is the bizarre appearance of a sine wave with sharp edges. The differential of wide, regular, monomorphic tachycardia is: VT or SVT with aberrancy, all +/- hyperkalemia (see diagrams below). Given the appearance of V1, hyperkalemia is diagnosed no matter what the underlying rhythm is. Anytime you find yourself confused as to what is the T-wave and what is the QRS complex, you are probably dealing with hyperkalemia.
Because hyperkalemia kills by either VT or VF arrest, or by bradycardia with PEA arrest, this should be assumed to be VT until proven otherwise.
We set the machine to synchronized cardioversion.
I of course should have known ahead of time that the default should be defibrillation, not synchronized, because (as I wrote) one cannot tell which is QRS and which is T-wave.
We caught it though:
The person manning the defibrillator was about to press shock when we remembered something very important to check in this scenario. If you do not have this ingrained into your mind as a reflex, you will miss this:
The defibrillator’s monitor was double-counting the heart rate. There were two monitor spikes for every one electrical cycle on the cardiac monitor. Upon closer inspection, the defibrillator’s monitor had assigned a spike for every large T wave as well as every QRS complex. Given the emergent nature of what was happening, I forgot to take a photo of this critical visual finding, but here is a non-hyperkalemia example:
Instead of running a 50/50 chance of shocking exactly on the T wave, we switched to unsynchronized cardioversion. A shock was delivered with the following change in rhythm:
The arterial line blood pressure immediately jumped from 80/40 with blunted monomorphic waveform to 140s/80s with bounding, normal-appearing waveform.
Over the next 30 minutes we noticed a repeating cycle, lasting about 15 minutes each time, characterized by widening of the QRS complex with progressively slower HR, followed by sudden onset of VT identical to the wide complex tachycardia above. Each time we shocked it back to sinus rhythm.
Meanwhile, we had been administering gram after gram of calcium chloride, as well as giving multiple amps of bicarb, multiple rounds of insulin and dextrose, and nebulizing albuterol continuously through the ventilator circuit. We should also have given 0.25mg intramuscular terbutaline, but we did not.
|Likely sinus rhythm with suppressed P-wave amplitude, but could also be junctional rhythm. The QRS complex is much more narrow than before, but still quite wide with “pulled-apart” morphology and T-wave abnormalities diagnostic of hyperkalemia. There are ST segment deviations appropriate for the QRS abnormalities and the clinical scenario.
Initial labs returned and confirmed critical hyperkalemia and new renal failure. Initial point of care chemistry simply read potassium greater than 8. 10 minutes later the lab chemistry revealed K = 9.7. Creatinine = 4.2 from baseline of 0.8 several weeks ago, with BUN = 109. Lactate = 11.2, troponin undetectable. pH < 6.88, bicarb 8.
(As a side note, if any reader can point us to literature detailing the potassium and troponin values in arrest, ideally stratified by down time, rhythm, cause of arrest, etc, please let us know. These questions frequently come up while caring for cardiac arrest patients, and I often find myself wondering what degree of elevations in troponin and/or potassium levels are attributable simply to cardiac arrest and down time)
After the 9th gram of calcium chloride (as well as the other therapies above), he finally stayed out of ventricular tachycardia and the QRS stopped re-widening. By this time my fellow resident Pete McKenna had already placed an internal jugular dialysis catheter.
|Slightly more narrow than prior.
His wife arrived and gave some very helpful history: He was a 53 year-old male with history of HTN, alcohol and substance abuse, and very recent new diagnosis of poorly differentiated metastatic cancer of unknown primary. He had just had his first oncology visit and had aggressive goals of care to treat his newly diagnosed cancer (no DNR/DNI). His wife reports coming home and finding the patient on the floor, responsive but lethargic, up until the point just prior to EMS arrival. She did not perform chest compressions prior to EMS arrival. Upon further history, it turns out the patient has been taking up to 24 pills of ibuprofen per day for the past month.
This is a reminder NOT to assume that arrest unwitnessed by EMS = huge downtime. This patient apparently was responsive until just minutes before EMS arrival. This is a very different patient with very different cause of arrest and chance of survival than the average unwitnessed PEA arrest. We would not have known this without detailed questioning of the family members.
|Even more narrow, with QRS approaching normal but the T-wave still showing peaking diagnostic of hyperkalemia.
He ended up requiring a total of 13 grams of CaCl in the ED to maintain rhythm and hemodynamics until our dialysis colleagues could start dialysis in the ED. No other cause of arrest was identified based on lab results or pan-CT scan. He made it to the ICU, however the patient unfortunately expired approximately 24 hours after ICU admission. While the outcome is disheartening, I think this case highlights some important management principles of a fairly common cardiac arrest phenotype. If we treat a number of patients with this cardiac arrest phenotype with this level of skill, some of them will survive with meaningful outcome.
1. Hyperkalemia is common, deadly, identifiable, and treatable. This is our territory in EM, as no cardiologist or nephrologist will be there to help us identify it on the ECG or start dialysis as quickly as these patients need it. We own the responsibility for this disease process, and no one else can do this better than us.
2. Wide, bizarre, brady, and blocks = hyperkalemia until proven otherwise. Sometimes it is sodium channel blockade, so emprically giving BOTH calcium and bicarbonate is sometimes worthwhile.
2. Anytime you find yourself confused about what is T-wave and what is QRS complex, assume two things: 1) this is hyperkalemia, and 2) the monitor will also be confused trying to choose what is T-wave vs. QRS complex when trying to synchronize
3. Really wide tachycardia = VT or SVT with conduction abnormality + hyperkalemia, until proven otherwise. No matter what you do about the VT vs. SVT part of the differential, do not fail to treat the hyperkalemia when it is a possibility.
See the following interesting case for another example:
4. Do not assume that unwitnessed arrest always means prolonged down time. Fight to get the most accurate information from EMS, family members, police, bystanders, etc.
5. No amount of calcium is too much in an unstable patient with hyperkalemia. We used 13 grams of calcium chloride in this case in the ED alone. Titrate first to ROSC, then to eu-rhythmia, then to narrow QRS complex (assuming a previously narrow QRS complex at baseline). The calcium will buy you approximately 15-20 minutes between episodes of deterioration, during which time you must exhaust all other methods of potassium redistribution and excretion.
In retrospect, we could have done even more in this case. We could have also considered terbutaline (start at 0.25 mg intramuscular) and/or Josh Farkas’ “nephron bomb”:
“Furosemide 80-160mg IV
+/- chlorothiazide 500-1000mg IV
+/- acetazolamide 250-500mg IV
+/- fludrocortisone 0.2mg PO”
Smith comment: I do not know of any data supporting this “nephron bomb,” so I neither endorse nor disapprove!
In 2018 there is generally a delay to onset of dialysis in most EDs. Somehow this must change, whether we decide to take this skill upon ourselves or convince our nephrology colleagues to set up more rapid access to their resources. These patients cannot afford a delay.
Think BRASH: do not underestimate the synergy of renal failure, AV blockade, shock, and hyperkalemia. Consider and correct each of these contributing factors in the treatment of hyperkalemia.