Be Skeptical… EM Myths and Their Evidence

It has become clear to me, at the end of a three year emergency medicine residency, that medicine is muddy, that there are few hard truths in our field…even in the age of evidence-based practice.   So many cases find us confronting critical patients with a woefully finite scope of knowledge, one that is evolving. That however, does not dissuade the teaching of medicine nor the conversations with consultants, pharmacists, surgeons, nursing staff, attendings, etc., nor the all-important internal discourse in our own minds from taking on an often absurd dogmatism. The following are just a few of the widely held and taught ideas whose time has passed….or never was.

Kayexalate (SPS)

What we’re told and what we’ve learned:

  1. Kayexalate safely eliminates potassium
  2. It is one of the three ways to decrease whole body K (along with diuretics and hemodialysis)
  3. (Hyperbole for effect) it is necessary for any patient being admitted for hyperkalemia

Standard of care for a hyperkalemic patient seems to be giving kayexalate.

The evidence behind its use:

  • There actually is no good literature based evidence for the effect of kayexalate on lowering serum potassium
  • Approved by FDA in 1958(probably fairly lax rules by today’s standards)
  • NEJM two studies in 1961
    • Scherr et al. 32 pts with hyperK and renal failure, various doses of SPS were given over various timeframes, with various daily dosing schedules and various courses of treatment(up to 280 days). Reduction of serum K by 0.4 in 66% pts after 24 hours. Not a RCT, no controls, all pts in the study were also on: low K diets, many received bicarb/insulin/glucose
    • Flinn et al, 10 pts, 5 kayexalate/sorbitol PO, 3 sorbitol, 2 kayexalate/sorbitol enema. All on low K diets and diuretics. 5 days QID dosing, steady decrease in K.
  • 1998 Study By nephrologist on 6 pts with ESRD
    • 5 regimens separated by 1 wk. SPS, SPS/sorbitol, SPS/phenolphthalein, phenolphthalein, placebo. Patients were treated for a week with each regimen, serum potassium was measured. Outcome: none were effective at reducing K.

The evidence for its harm:

  • By 2005, 35 case reports of colonic/bowel necrosis reported
  • At risk groups: renal failure, transplant, burn patients
  • Surgery, Lillemoe et al. 1987 Case series
    • 5 post op pts with uremia/HyperK suffered “catastrophic colonic necrosis”, 4/5 died (all with sorbitol)
  • Southern Medical Journal, McGowen 2009
    • Case Series over 9 years, at a single institution, 11 cases of bowel necrosis
  • In many of these cases, sorbitol was thought to be the culprit, and in light of this information, the FDA recommended only using 33% sorbitol formulation(2009), then later the manufacturer completely removed sorbitol from its kayexalate formulations.
  • Harel 2013. Systematic Review of case reports and series of complications of SPS
    • GI complications in patients taking SPS: 58 total, 41 with sorbitol, 17 without. 19 fatalities. Most common complication: transmural necrosis

The takeaway message:
Kayexalate, at best, is an ineffectual drug that theoretically reduces serum potassium levels as a cation exchange resin. Kayexalate, at worst, is an ineffectual drug that may case colonic necrosis. “First do no harm…”, advocate for your patients. There is no good evidence to support the use of a drug that has been associated with severe complications, so why do we continue to use it?

There is a very good discussion of this topic in a 2013 EMRAP episode:

Digoxin, Calcium, and Stone Heart

What we’re told and what we’ve learned: If you give calcium to a digoxin intoxicated patient, her heart will TURN TO STONE, or rather, precipitate an irreversible contractile state…also, malignant dysrhythmias will ensue.

The theory:

  • Calcium may lead to an irreversible non-contractile state, due to impaired diastolic relaxation from calcium-troponin C binding

The evidence:

  • In animal models, administration of calcium to digoxin intoxicated subjects
  • Case reports in humans that argue for a possible temporal relationship between calcium administration and death

No, really, the evidence:

  • In the animal models showing a causal relationship between negative effects of calcium and digoxin toxicity, serum calcium levels were severely artificially elevated, often >14 mM/L, before digoxin was given.
  • Further animal studies were not able to reproduce toxic effects of calcium administration to digoxin intoxicated canines
  • The case report evidence in human patients is scant, amounting to only five case reports, all coming from the 1930s-1950s.
    • These case reports to not prove causality, but rather softly imply a relationship between calcium administration and death in the setting of digoxin toxicity.
    • There was no way to measure digoxin levels at the time of these case reports, so all digoxin toxicity was presumed, though no symptomatology of digoxin toxicity was described in these reports.
    • It is highly likely that these patients were very ill at the time of calcium administration, with CHF comorbidity and hyperkalemia at least.

The evidence on the other side:

  • Levine et al. published in the Journal of Emergency Medicine in 2011: The Effects of IV Calcium in Patients with Digoxin Toxicity
  • The study was a retrospective chart review from a single institution
  • It looked at all patients from January 1, 1989 – May 31, 2005 with a serum digoxin level > 2.0 ng/dl or with a diagnosis of digoxin toxicity,
  • The study was looking for these outcomes: potentially fatal dysrhythmias(i.e. ventricular fibrillation, sustained ventricular tachycardia, Mobitz II second-degree heart block, complete heart block, or asystole) at 1 hour after IV calcium administration and at 4 hours after IV calcium administration, as well as overall mortality
  • There were a total of 159 patients found, of these, 23 patients were given IV calcium and 136 were not given calcium.
  • No dysrhythmia occurred within 4 hours of calcium administration
  • 5/23 patients given IV calcium died (22%) versus 27/136 (20%) of those who did not receive calcium

The article:

The takeaway message:
In a case of a critically ill patient with hyperkalemia, or EKG findings concerning for hyperkalemia, where seconds matter, giving IV calcium without any digoxin toxicity angst rattling around in your head, is probably a very safe thing to do.

Digoxin toxicity being a contraindication for administration of IV calcium in the setting of hyperkalemia is likely a false concern.

If a patient is dig toxic and hyperkalemic, there is an antidote: Digibind.

Dopamine in Cardiogenic Shock

What we’re told and what we’ve learned: Dopamine is the first line pressor for cardiogenic shock with profound hypotension.

The evidence for dopamine:

  • “That’s what we do”
  • RCTs lacking
  • De facto standard of care at most institutions, trickled down from cardiology

However, Circulation 2008:

  • On cardiogenic shock: “Data on comparison of vasopressors are scant. The American College of Cardiology/American Heart Association (ACC/AHA) guidelines recommend norepinephrine for more severe hypotension because of its high potency. Although both dopamine and NE have inotropic properties, dobutamine is often needed in addition.”

The evidence against dopamine:

  • NEJM 2010 De Backer et al
    • Comparison of Dopamine and Norepinephrine in the Treatment of Shock
    • Prospective RCT pitting dopamine vs norepinephrine in shock, all-comers
    • 1679 pts, 858 randomized to dopamine, 821 randomized to NE
    • No significant between-group difference in the rate of death at 28 days (52.5% in dopamine group, 48.5% in NE group; odds ratio with dopamine, 1.17; P=0.10)
    • More arrhythmic events among the patients treated with dopamine than with NE(207 events [24.1%] vs. 102 events [12.4%], P<0.001)
  • Subgroup analysis showed that dopamine, as compared with norepinephrine, was associated with an increased rate of death at 28 days among the 280 patients with cardiogenic shock but not among the 1044 patients with septic shock or the 263 with hypovolemic shock

The article:

The takeaway message:

  • Norepinephrine is the first line pressor of choice in cardiogenic shock
  • From Current Opinion Cardiology, May  2014
    • “Inotropes and vasopressors are widely used to improve hemodynamics acutely; however, reliable information regarding comparative efficacy of individual agents is lacking. A subanalysis of a prospective randomized trial suggested that norepinephrine may be preferredover dopamine in patients with cardiogenic shock.”

Cross Reactivity of Penicillins

What were told and what we’ve learned: There is an approximate 10% cross reactivity for penicillin allergic patients to cephalosporins.

What investigation reveals: The actual cross-reactivity rate of allergy is 1-2.5%.

The data: A recent article published in 2012:

  • The use of cephalosporins in penicillin allergic patients: A literature review. J Emerg Med 2012
    • Review of 27 articles/studies dating back to 1950
    • Overall cross-reactivity rate between cephalosporins and penicillins in patients reporting a penicillin allergy = 1% 
    • Overall cross-reactivity rate in patients with a confirmed penicillin allergy = 2.5%
    • The true incidence of an allergy to penicillin in patients believed to have such allergy is <10%

Why the confusion?

  • The theory accounting for the cross reactivity of allergy between penicillins and cephalosporins is that the beta-lactam rings are similar.
  • When this 10% statistic was being tallied and calculated, penicillins were being co-produced with cephalosporins using the same fungus. When this was occurring, there was a high risk of contamination between the two classes of antibiotic.
  • More recent studies have determined the actual risk of cross-reactivity relates more to a R-1 side chain similarity and probably not the beta-lactam ring.

The takeaway message:

  • From J Emergency Medicine 2012: “Conclusions: Although a myth persists that approximately 10% of patients with a history of penicillin allergy will have an allergic reaction if given a cephalosporin, the overall cross-reactivity rate is approximately 1% when using first-generation cephalosporins or cephalosporins with similar R1 side chains…. For penicillin-allergic patients, the use of third- or fourth-generation cephalosporins or cephalosporins with dissimilar side chains than the offending penicillin carries a negligible risk of cross allergy.”
  • There is negligible cross-reactivity between most 1st and 2nd generation cephalosporins and penicillins

The article:

IVFs and Cerebral Edema in Pediatric DKA

What we’re told and what we’ve learned:

  • Be very careful with fluid resuscitation in pediatric patients with DKA lest you cause permanent brain damage or death even
  • Journal of Pediatric Critical Care, 2008, Cerebral edema in dka
    • Review article
    • Fluid Administration. It seems prudentto avoid excessive amounts of fluid, fluid given too rapidly, and use of hypotonic fluid. Most authors do not recommend that a bolus of fluid be given unless there is evidence of cardiovascular compromise as demonstrated by extreme tachycardia, hypotension, cold extremities, and/or anuria.”

The evidence for:

  • >20 retrospective studies.
  • No prospective trials
  • No causal association between IVFs and cerebral edema has been proven
  • J Ped Crit Care 2008
    • “Therefore, risks are also associated with therapy with insulin, which lowers blood glucose and stimulates the sodium hydrogen exchange mechanism, increasing intracellular sodium concentration and fluid (especially hypotonic fluid) given to repair dehydration. Both of these interventions can cause a rapid decrease in serum osmolality and favor movement of water into brain cells, although the causal relation of fluid administration to CEDKA is debated”

The evidence against:

  • No one knows the mechanism of cerebral edema in DKA, in fact, the opposite is argued, that it is cytogenic cerebral edema caused by breakdown from cerebral hypoperfusion and that fluid resuscitation may prevent its occurrence.
  • Literature is full of cases of MRI/CT done prior to treatment of DKA indicating cerebral edema existing
  • NEJM 2001, Glasner, Risk factors for development of DKA
    • Retrospective study, encompassing 15 yrs
    • 7000 pediatric pts, ages 4-13, 61 cases of CE
    • No correlation between rate of fluid admin or tonicity of fluid and development of CE
  • Only Prospective study: Pediatrics 2013, Glasner, Subclinical Cerebral Edema in Children with DKA…
    • 18 patients (ages 8-18) randomized to 20 cc/kg initial + rapid replacement of estimated fluid deficit (8 patients) vs 10cc/kg + slow replacement of estimated fluid deficit over 2 days (10 patients)
    • Same insulin gtt started
    • MRI done at three different times, at 3-6 hours, 9-12 hours, and after resolution
    • No cases of clinical CE
    • Looking for changes in apparent diffusion coefficient measured(precursor to cerebral edema)
    • No difference in groups

The takeaway message:

  • There is absolutely no good evidence that generous IVF resuscitation in pediatric DKA causes cerebral edema, only one of many theories as to the cause of cerebral edema
  • Treat the sick/dehydrated patient with IVFs aggressively. If you are treating a child that is shocky and critically ill in DKA, don’t skimp on IVFs because in the back of your mind you are hearing the unsubstantiated ravings of a pediatric endocrinologist
  • Document a good initial neuro exam, it may be that DKA itself causes that rare complication of cerebral edema, and it may be present upon initial presentation
  • Careful with age < 5 as these studies did not include children under the age of five
  • FLUID Trial coming…large prospective trial pitting aggressive vs non-aggressive IVF resuscitation in pediatric DKA patients

The articles:

An ABG Is a Necessary Laboratory Study

What we’re told and what we’ve learned: You don’t know a patient’s pH, pCO2, HCO3, base excess, pO2 until you have an ABG.

What the evidence shows: A VBG will suffice in the overwhelming majority of cases

The evidence for VBG:

  • pH: good correlation, mean difference 0.03-0.04
  • HCO3: good correlation, mean difference 0.52-0.9
  • Base excess: correlation mean difference 0.089 mmol/L
  • pCO2: good correlation mean difference -4
  • 100% sensitive in detecting arterial hypercarbia in COPD
    • Venous vs arterial blood gases in the assessment of patients presenting with an exacerbation of chronic obstructive pulmonary disease. Am J Emerg Med. 2012
  • pH assessment in DKA, mean difference -0.015 +/- 0.006 units
    • ABG results rarelyinfluence emergency physician management of patients with suspected diabetic ketoacidosis.Acad Emerg Med 2003, Ma et al
    • 200 cases of DKA, ABG and VBG drawn simultaneously
    • Result of ABG changed treatment/disposition in 2/200 pts
  • But what about the pO2?
    • Remember the pulse oximeter and the oxyhemoglobin saturation curve? Gives you a pretty great guess at pO2

The evidence against ABG:

  • pH, PCO2 (if normocapneic), HCO3 and base excess from a VBG are adequate for clinical decision making
  • SpO2 is usually sufficient for clinical decision making unless pulse oximetry is unreliable for other reasons
  • increased risk of bleeding and hematoma
  • risk of pseudo aneurysm and AV fistula
  • Infection
  • nerve injury
  • digital ischemia
  • injury to staff
  • delays in care
  • serial exams may be needed
  • venous sampling may better represent the tissue milieu

The takeaway message:

  • Do more VBGs, do less ABGs , there are very few cases when an ABG is clinically necessary
  • Use ETCO2, great non-invasive surrogate for VBG and ABG in cases of hypercapnia
  • Advocate for the human being that is your patient, really silly to jab the radial artery of every patient in respiratory distress, intubated, hypoxic, altered, etc.
  • If you have an art line, go nuts with ABGs
  • ABG If you can’t trust the sats, i.e. not reading due to shock/cold extremities, carbon monoxide poisoning, methemoglobinemia
  • ABG may be necessary to find true extent of hypercarbia as VBG CO2 has been shown to be less correlative at higher pCO2.       However, knowing whether the pCO2 is 90 or 100 in that altered patient in respiratory distress will likely not change your management. (more of an academic exercise in patient harming futility)

The articles:

Checking Troponin in SVT

What we’re told and what we’ve learned: Send troponins in SVT/chest discomfort

What the evidence shows: Routine testing of pts in SVT with troponin is not necessary

The data:

  • 12 – 48% of patients have elevated troponins after SVT
  • Having a known history of CAD is more likely to lead to troponin elevation than not having a history of CAD (62% vs 43%)
  • There is no difference in the diagnosis of CAD compared to patients with negative troponins
  • Can J Cardio, 2011, Yedder et al. Retrospective review
    • 73 pts with SVT, troponin checked
    • 23+, all underwent stress or LHC, 2 +stress got LHC
    • All negative
    • Patients with +trop had significantly higher maximum HRs (190.8 versus 170.3 beats per minute, P = .008). A correlation was found between the maximal heart rate during SVT and the level of troponin elevation (r = 0.637, P = .001)

Few more papers…

  • Kanjwal et al (2008), Case Series (7 patients)
    • All with NORMAL coronary arteries
  • Miranda et al (2006), Case Study
    • NORMAL coronary arteries
  • Yeo et al (2006), Case series (3 patients)
    • All with NORMAL coronary arteries
  • Redfearn et al (2005), Retrospective Cohort (3 patients)
    • All with NORMAL coronary arteries
  • Zellweger et al (2003), Case Series (4 patients)
    • 2 patients with NORMAL coronary arteries on angiography and 2 patients with normal myocardial perfusion scans
  • Bakshi et al (2002), Prospective Cohort (3 patients)
    • All with NORMAL coronary arteries

Chow et al, Medicine 2010, Prognostic significance of trop in SVT

  • Group of patients with SVT who had troponin testing done followed out to one year
  • Patients with more comorbidities(risk factors for ACS) are more likely to have elevated troponins with SVT
  • In comparing elevated versus non-elevated troponin I levels post-SVT, the incidence of adverse outcomes was:
    • Cardiac rehospitalization: 37.9% versus 6.1% (afib and chf)
    • Myocardial infarction: 6.9% versus 0% (Not Statistically Significant)
    • Death: 20.7% versus 10.2% (Not Statistically Significant)
    • So..less healthy people are less healthy

The takeaway message:

  • In young patient with known SVT, don’t check troponin unless HPI dictates, leads to possible iatrogenic cascade of misadventure and wasting of resources
  • If patient is older, with many co-morbidities, remarkable chest pain, etc. send a troponin on a case by case basis
  • Remember, SVT will often have rate related ST changes that can persist 2-3 hours after cardioversion

Check Coagulation Studies Before Paracentesis

What we’re told and what we’ve learned: It is prudent to have an INR and platelets prior to paracentesis procedure

The evidence against the necessity of coags:

  • The Use of Paracentesis in the Assessment of the Patient With Ascites, Wilkerson, Annals of Emerg Med, 2009
  • Review of articles, includes 2 prospective trials with coags/platelets drawn before procedure
  • 310 pts, No instances of significant bleeding, 2 minor cases

The takeaway message:

  • Conclusions: In a patient in whom spontaneous bacterial peritonitis is suspected, paracentesis should be performed. Coagulation studies are likely not required before performance of the procedure.”
  • “The procedure of paracentesis has been considered by some to be a potential barrier for the emergency physician’s rapid diagnosis of spontaneous bacterial peritonitis. Concerns over the patient’s coagulopathy…These issues should now be considered assuaged by this evidence-based review, allowing emergency physicians to comfortably and expeditiously use paracentesis to investigate spontaneous bacterial peritonitis in any decompensated patient with ascites.” Annals of Emerg Med, 2009

The article:

References / Further Reading


  • Scherr L, Ogden DA, Mead AW, et al. Management of hyperkalemia with a cation-exchange resin. N Engl J Med 264: 115-9, 1961.
  • Flinn RB, Merrill JP, Welzan WR. Treatment of the oliguric patient with a new sodium ion exchange resin and sorbitol: A preliminary report. N Engl J Med 264: 111-5, 1961.
  • Lillemoe, K. D., et al. “Intestinal necrosis due to sodium polystyrene (Kayexalate) in sorbitol enemas: clinical and experimental support for the hypothesis.” Surgery 101.3 (1987): 267.
  • McGowan, C. E., et al. “Intestinal necrosis due to sodium polystyrene sulfonate (Kayexalate) in sorbitol.” Southern medical journal 102.5 (2009): 493.
  • Harel, Ziv, et al. “Gastrointestinal adverse events with sodium polystyrene sulfonate (Kayexalate) use: a systematic review.” The American journal of medicine 126.3 (2013): 264-e9.


Stone Heart


Dopamine in Cardiogenic Shock

  • De Backer, Daniel, et al. “Comparison of dopamine and norepinephrine in the treatment of shock.” New England Journal of Medicine 362.9 (2010): 779-789.
  • Reynolds, Harmony R., and Judith S. Hochman. “Cardiogenic shock current concepts and improving outcomes.” Circulation 117.5 (2008): 686-697.
  • Nativi-Nicolau, Jose, et al. “Pharmacologic therapies for acute cardiogenic shock.” Current opinion in cardiology 29.3 (2014): 250-257.


Cross Reactivity of Penicillins


Cerebral Edema in DKA

  • Levin, Daniel L. “Cerebral edema in diabetic ketoacidosis.” Pediatric Critical Care Medicine 9.3 (2008): 320-329.
  • Glaser, Nicole, et al. “Risk factors for cerebral edema in children with diabetic ketoacidosis.” New England Journal of Medicine 344.4 (2001): 264-269.
  • Glaser, Nicole S., et al. “Subclinical cerebral edema in children with diabetic ketoacidosis randomized to 2 different rehydration protocols.” Pediatrics 131.1 (2013): e73-e80.



  • Middleton P, Kelly AM, Brown J, Robertson M. Agreement between arterial and central venous values for pH, bicarbonate, base excess, and lactate. Emerg Med J. 2006 Aug;23(8):622-4.
  • Ma OJ, Rush MD, Godfrey MM, Gaddis G. Arterial blood gas results rarely influence emergency physician management of patients with suspected diabetic ketoacidosis. Acad Emerg Med. 2003 Aug;10(8):836-41.
  • McCanny P, Bennett K, Staunton P, McMahon G. Venous vs arterial blood gases in the assessment of patients presenting with an exacerbation of chronic obstructive pulmonary disease. Am J Emerg Med. 2012 Jul;30(6):896-900.
  • Kelly AM, McAlpine R, Kyle E. Venous pH can safely replace arterial pH in the initial evaluation of patients in the emergency department. Emerg Med J. 2001 Sep;18(5):340-2.


Troponin in SVT

  • R.N. Bukkapatnam, M. Robinson, S. Turnipseed, D. Tancredi, E. Amsterdam, and U.N. Srivatsa, “Relationship of myocardial ischemia and injury to coronary artery disease in patients with supraventricular tachycardia.”, The American journal of cardiology, 2010.
  • M. Dorenkamp, M. Zabel, and C. Sticherling, “Role of coronary angiography before radiofrequency ablation in patients presenting with paroxysmal supraventricular tachycardia.”, Journal of cardiovascular pharmacology and therapeutics, 2007.
  • G.V. Chow, G.A. Hirsch, D.D. Spragg, J.X. Cai, A. Cheng, R.C. Ziegelstein, and J.E. Marine, “Prognostic significance of cardiac troponin I levels in hospitalized patients presenting with supraventricular tachycardia.”, Medicine, 2010.
  • Ben Yedder, Naïm, Jean Francois Roux, and Felix Ayala Paredes. “Troponin elevation in supraventricular tachycardia: primary dependence on heart rate.” Canadian Journal of Cardiology 27.1 (2011): 105-109.


Coags in Paracentesis

  • Wilkerson, R., & Sinert, R. (2009). The Use of Paracentesis in the Assessment of the Patient With Ascites. Annals of Emergency Medicine, 54 (3), 465-468 DOI: 10.1016/j.annemergmed.2008.09.005
Ryan Dietert, MD is an EM senior resident at UTSW/Parkland. This article was edited by Alex Koyfman, MD.

3 thoughts on “Be Skeptical… EM Myths and Their Evidence”

  1. Great post on medical myths and, as Cliff Reid calls it, dogmalysis.

    I would add some more on calcium in digoxin poisoning. The issue is slightly more complicated. Even if you don’t believe the stone-heart theory (which I do not) there’s another reason not to bother with calcium; it doesn’t work to stabilize the myocytes in this case. Bob Hoffman has a nice discussion of this here (

    Basically, in dig toxicity, intracellular calcium levels are high and physiologically, it’s unlikely that additional calcium will have any effect. Additionally, Hoffman argues that patients with dig toxicity are unlikely to have EKG effects from the hyperkalemia due to the mechanism of poisoning. I’m not tox expert so I encourage emDocs followers to check out the FOAM article from Hoffman. They key is to not wast time on this intervention. Digibind is the way to go.

  2. Great post on medical myths and, as Cliff Reid calls it, dogmalysis.

    I would add some more on calcium in digoxin poisoning. The issue is slightly more complicated. Even if you don’t believe the stone-heart theory (which I do not) there’s another reason not to bother with calcium; it doesn’t work to stabilize the myocytes in this case. Bob Hoffman has a nice discussion of this here (

    Basically, in dig toxicity, intracellular calcium levels are high and physiologically, it’s unlikely that additional calcium will have any effect. Additionally, Hoffman argues that patients with dig toxicity are unlikely to have EKG effects from the hyperkalemia due to the mechanism of poisoning. I’m not tox expert so I encourage emDocs followers to check out the FOAM article from Hoffman. They key is to not wast time on this intervention. Digibind is the way to go.

Leave a Reply

Your email address will not be published. Required fields are marked *