ECG Changes of Hyperkalemia

Hyperkalemia is an electrolyte abnormality seen in the emergency department as well as in hospitalized patients and it can be associated with adverse clinical outcomes and death if not treated appropriately. It is important to remember that the electrophysiologic effects of hyperkalemia are directly proportional to both the absolute plasma potassium and its rate of rise. However, neither the ECG nor the plasma potassium alone are an adequate index of the severity of hyperkalemia, and therefore providers should have a low threshold to initiate therapy. Classic teaching of the chronological ECG changes of hyperkalemia include:

  1. Peaked T waves
  2. Prolongation of PR interval
  3. Widening QRS Complex
  4. Loss of P wave
  5. “Sine Wave”
  6. Asystole

Where did the evidence for order of ECG changes of hyperkalemia come from?

The order of ECG changes of hyperkalemia have been defined in the experimental setting, but no uniform order has been documented in animal models [1]. The relationship between serum potassium and cardiac manifestations is even less clear in the clinical setting [2].

How good are ECGs at predicting hyperkalemia? [3] 

  • 87 patients admitted from the ED with hyperkalemia
  • ECG read by 2 EM physicians blinded to potassium level
  • Sensitivity: 34 – 43%
  • Specificity: 85 – 86%
  • When only patients with potassium > 6.5 mmol/L were analyzed: Sensitivity 55 – 62%
  • Conclusion: Based on this study the ECG has a poor correlation with detection of hyperkalemia

Any studies looking at the frequency of ECG changes with hyperkalemia? [4]

What is the treatment regimen for hyperkalemia? [5]

Hyperkalemia Treatment

Additional Treatment Pearls:

  1. 1 g Calcium Chloride (CaCl) = 3 g Calcium Gluconate: Bryan Hayes (@PharmERToxGuy) had a great post on Academic Life in EM on myths associated with different calcium regimens.
  2. Hypertonic (3%) Saline: Should be restricted to patients with hyponatremia and concurrent hyperkalemia.  Its effects have not been established in eunatremic patients.
  3. Insulin: 10 Units of regular insulin + dextrose (25 g as 50% solution) lowers plasma potassium by about 0.6 mmol/L. Be careful as a single bolus of 25 g of dextrose has been shown to be inadequate ion prevention of hypoglycemia at 60 minutes.
  4. Beta Agonists: 10 mg  and 20 mg of inhaled albuterol decreased serum potassium by 0.6 mmol/L and 1 mmol/L, respectively. Mild tachycardia is most common side effect.
  5. Bicarbonate: Bolus injection of sodium bicarbonate has not been studied.  The dogma of bolus bicarbonate was derived from studies using a prolonged (4 – 6 hr) infusion of bicarbonate which does not reduce plasma potassium in patients with dialysis-dependent renal failure. Infusion sodium bicarbonate over 4 – 6 hours may have some benefit in excretion of serum potassium in the setting of metabolic acidosis.
  6. Exchange Resin: The onset of action is very long (2 hours to onset, and 6 ours to maximum effect) and there have been case reports of patients developing colonic necrosis. Also there have been studies that show no statistical benefit at 4, 8, and 12 hours [6].
  7. Hemodialysis: Definitive treatment of hyperkalemia. Watch out for rebound hyperkalemia which can occur in 30% of patients an hour after dialysis.

Clinical Bottom Line: Neither the ECG nor the plasma potassium alone are an adequate index of the severity of hyperkalemia, and providers should have a low threshold to initiate cardiac membrane stabilization in the setting of hyperkalemia and no ECG changes.


  1. Ettinger PO et al. Hyperkalemia, Cardiac Conduction, and the Electrocardiogram: A Review. Am Heart J 1974. PMID: 4604546
  2. Acker CG et al. Hyperkalemia in Hospitalized Patients: Causes, Adequacy of Treatment, and Results of an Attempt to Improve Physician Compliance with Published Therapy Guidelines. Arch Intern Med 1998. PMID: 9570179
  3. Wrenn KD et al. The Ability of Physicians to Predict Hyperkalemia from the ECG. Ann Emerg Med 1991. PMID: 1952310
  4. Montague BT et al. Retrospective Review of the Frequency of EC Changes in Hyperkalemia. clinical J Am Soc Nephrol 2008. PMID: 18235147
  5. Weisberg LS. Management of Sever Hyperkalemia. Crit Care Med 2008. PMID: 18936701
  6. Gruy-Kapral C et al. Effect of Single Dose Resin-cathartic therapy on Serum Potassium Concentration in Patients with End-Stage Renal Disease. J Am Soc Nephrol 198. PMID: 9773794

For More on This Topic Checkout:

Cite this article as: Salim Rezaie, "ECG Changes of Hyperkalemia", REBEL EM blog, April 17, 2014. Available at:
The following two tabs change content below.

Salim Rezaie

Emergency Physician at Greater San Antonio Emergency Physicians (GSEP)
Creator & Founder of REBEL EM

Like this article?

Share on facebook
Share on Facebook
Share on twitter
Share on Twitter
Share on linkedin
Share on Linkdin
Share on email
Share via Email

Want to support rebelem?

19 thoughts on “ECG Changes of Hyperkalemia”

  1. Salim – nice succinct post. Hyperkalemia is common and life-threatening and so it’s vital for us to immediately recognize it and treat the patient rapidly. Treatment of hyperkalemia is one of those “must memorize” as opposed to “must know where to find” algorithms.

    I do disagree about the utility of resins and have stopped using them. The evidence for their utility is minimal and it’s not recommended by most nephrologists. We did a longer post on this on EM Lyceum here

    I haven’t tried the hypertonic saline but will have to give in a shot next time I’ve got hyperk and hypona. Thanks!

    • Thanks Anand,
      I 100% agree about the resins. In the acute setting it is of little use with greater than 2 hours to the onset of effect and 4 – 6 hours to maximal onset of effect. It’s funny it still gets listed in textbooks, but I find minimal use for it in the ED as well.


  2. Guys,

    Thanks for the great post.

    I love this topic, and am constantly forwarding articles to people…

    What I would love to see added to these discussions is the expected effect. That is how much shifting and/or excreting of K+ do you expect.

    In the ED it seems we often ‘need everything we can get’ (>6.0) or we probably don’t need to do anything(<5.5)…

    What do you tell your residents/colleagues about how much bang for the buck you get with each treatment?

    thanks again.


    • Hello Steve,
      TY for reading. As best I can tell….

      1. Regular insulin 10U IV + dextrose (25 g as 50% solution) lowers plasma K+ by about 0.6 mmol/L
      2. 10 mg and 20 mg of inhaled albuterol lowers serum K+ by 0.6 mmol/L and 1 mmol/L, respectively
      3. In the setting of metabolic acidosis, a 4 hour infusion of 150mmol/L Bicarbonate at various rates lowers serum K+ by 0.6 mmol/L. A bicarbonate infusion administered over 4 –6 hrs at a rate designed to alkalinize the urine may enhance urinary K excretion.
      4. Loop Diuretics depend on renal function. The worse off the kidney the less likely the lasix will be effective. That being said…the truth is not sure anyone knows exactly
      5. Exchange Resins: In the acute setting these don’t really help and there are studies that show that at 4, 8, and 12 hours there is no effect on the serum K+, so I don’t use this. Also there are case reports of colonic necrosis documented as well.

      Hope this helps.


  3. Hi Salim,
    I enjoy teaching this topic to Ed nurses as it is one of those scenarios that the nurse should not be messing around and they need to know the order of treatment – Stabilization (if warranted), shifting and excreting.
    I am curious to know if one does throw the ‘kitchen sink’ at them for a significantly elevated level, do all these efforts work cummulatively or is the additive effect not as profound as an individual treatment. I’m not sure if studies have isolated treatment effect versus cummulative effect and if their is a combination that is better than others.
    Thanks for your blog!

    • Hello Brian,
      TY for the support of the site and reading the post. I am not aware of a study that looks at comparisons of different treatments as additive or cumulative with different combinations of treatments, but clinically if there are ECG changes I think you do everything you can to avoid the dreaded sine wave or vfib arrest….i.e. throwing the kitchen sink. More importantly, we don’t truly need to know the exact decrease in potassium, but if whatever the decrease was got rid of the ECG changes which is more important IMHO. Thank you again for reading and your question. If I find anything regarding this I will add it to the post.


  4. Salim – You have definitely gathered together some nice information about HyperKalemia. That said – I do NOT agree with at least 2 of the two generalized conclusions you draw:

    i) Your Statement – “The ECG is not a sensitive method to detect hyperkalemia”. There is a LOT we don’t know re the details of this study that make the “conclusion” NOT supported (I can’t tell if this is the authors’ conclusion – your conclusion – or both you and the authors’ conclusion). ECG changes of hyperkalemia are superimposed on whatever the baseline ECG shows. “Sensitivity” will obviously be far less IF the baseline ECG is abnormal. We have NO idea what the baseline ECG in these 87 cases looked like? We have no idea of the skill of these 2 ED physicians in ECG interperation (Maybe they are among the world’s best in ECG interpretation – maybe not ….). We have no idea of the NUMBER of patients at each serum potassium level (I would REALLY like to see the actual ECGs of those patients with K+ > 6.5mEq/L who made up the group with such poor sensitivity despite this high a level. A far BETTER way of expressing the concept you are after – is the “correlation” between relative amount of serum K+ elevation and ECG findings. While clearly NOT perfect – that overall correlation of the ECG to the serum K+ value is really PRETTY GOOD from a qualitative viewpoint (ie, as clinical assistance in evaluating and treating patients, esp. in the right clinical setting).

    ii) Your Clinical “Bottom Line” – which states, “… providers should have a low threshold to initiate therapy in the setting of hyperkalemia and no ECG changes”. Hard to know what you mean by “no ECG changes” – but while emergency treatments for severe hyperkalemia are generally not toxic – it is hard to believe that you would treat a patient with acute renal failure with full array of treatment measures IF serum K+ is less than 5.0 mEq/L and the ECG is stone cold normal (= narrow QRS, normal sinus rhythm with good amplitude P wave, and totally normal ST-T waves without slightest hint of T wave peaking). In my opinion – to advise “blanket treatment” (as the wording of your Clinical “Bottom Line” implies) is not a correct message. ON the contrary – the combination of: i) serum K+ value – PLUS ii) one or more ECGs on the patient PLUS – the clinical context in which you assess i) + ii) are EXCELLENT for providing indication of whether or not emergency treatment measures for hyperkalemia are indicated – realizing that at time close follow-up and serial values of assessment tests will be needed to provide the full answer.

    THANK YOU for your receptivity of my constructive comments. I just think meticulous attention to wording and implications of diagnostic measures and therapeutic interventions is essential when dealing with this particular topic that is so important to emergency providers.

    • Hello Ken,
      Of course you know I don’t mind feedback and thoughts on anything on this site. I enjoy your thoughtful comments and knowledge on here. It not only makes me a better physician but everyone else who comes onto the site. Ok, I will try and answer these one at a time…

      1. For the sensitivity comment…point taken. I 100% agree with what you are saying and do consider that a limitation of the study. We don’t know what “normal” and “abnormal” really mean in these 87 cases. It is well known that if the baseline ECG is “abnormal” at baseline, some changes may not be as evident…so to make the blanket statement, you would actually need to see the 87 ECGs, which you are correct are not shown in the study. I like your wording of correlation as opposed to sensitivity. I will change this in the post ASAP. As for the conclusion made…it was the conclusion of the author.

      2. I will try and sharpen up the wording on this as well. My point was even if there are no typical findings of hyperkalemia….peaked t-waves, widening of QRS…one should strongly consider initiating cardiac membrane stabilization. what does one amp of calcium gluconate really hurt? Minimal side effects and not expensive. So by treatment I mean cardiac membrane stabilization which I will also sharpen up the wording for. TY for bringing that up.

      Please always feel free to continue putting comments on the posts. Without these, we cannot continue the conversation of improving our patient care. TY.


  5. Hi Salim – THANK YOU (as always!) for your receptivity to my feedback. Sharpening up your wording may be all that is needed. I think we generally agree on the concepts. I’m fine with presumptive treatment until serum values come back when there is high clinical suspicion of hyperkalemia. Keep up the good work! – 🙂

  6. This has been an interesting discussion & my sincere gratitude to all, especially to Salim.

    I have a query & would be happy if someone could clarify. We know that hyperkalemia has it’s major adverse effects on the heart. We also know that in patients with CKD, ECG changes of hyperkalemia start appearing at higher s. potassium levels.

    Considering the above, are there studies which show the relationship between s. potassium levels, the presence of ECG changes & cardiotoxicity. Putting it another way, say we have 2 patients, both with a s. potassium of 6.5 mEq/l. One patient does not have any ECG changes while the other has tenting of the T waves. Would the first patient be less prone to develop cardiotoxicity? Is it possible that the first patient is able to tolerate higher levels of s. potassium for some reason or the other? This information would be interesting given that the major toxicity of hyperkalemia is on the heart. Therefore if a patient has a s. potassium of 6.5 but the ECG is normal, can we be more cautious before initiating treatment for the hyperkalemia?


Leave a Comment

Time limit is exhausted. Please reload CAPTCHA.