SCA Bulletin Home

Print Issue as PDF

See Full Table of Contents

In this Edition

2017 SCA Annual Meeting

SCA News

Literature Review

SCA Newsletter Committee

SCA Foundation


SCA Quick Links

Literature Review

Potassium and Magnesium Supplementation Do Not Protest Against Atrial Fibrillation After Cardiac Operation: A Time-Matched Analysis

Lancaster TS, Schill MR, Greenberg JW, et al. Ann Thorac Surg. 2016;102(4):1181-1188.

Reviewers: Melina Eshelman, MD1; Antonio Hernandez Conte, MD MBA2

  1. Department of Anesthesiology, University of California—Irvine, Orange, CA
  2. Department of Anesthesiology, Kaiser Permanente Los Angeles Medical Center, Los Angeles, CA

Hypokalemia and hypomagnesemia are thought to increase risk of postoperative atrial fibrillation (POAF) after cardiac operation, although this relationship has not been well-characterized. Several studies have shown an association between low potassium levels and ventricular but not atrial arrhythmias. Low magnesium levels also have been associated with an increase or no change in POAF risk. Despite conflicting data, electrolyte supplementation is commonly used for POAF prevention and treatment. The purpose of this study was to characterize the natural time course of plasma potassium and magnesium levels after cardiac surgery and to evaluate the effect of electrolyte levels and supplementation on the occurrence of POAF. 


This study was a retrospective review of all adult patients undergoing coronary artery bypass grafting (CABG), valve operation, or CABG plus valve operation at one medical center over a period of 5 years. Patients with a preoperative history of atrial fibrillation (AF) or atrial flutter were excluded. The study cohort was divided into 2 groups based on the occurrence of new-onset POAF, which was defined as new AF or atrial flutter or at least 30 seconds duration occurring between the times of operating room exit and hospital discharge. The control group consisted of patients with no AF documented for the duration of their postoperative stay. 

The collection times and results for one baseline and all postoperative plasma potassium, magnesium, and creatinine levels were obtained, as well as the administration times and dosing of all potassium and magnesium supplementations given during hospital admission. A standardized protocol was used for repletion of potassium but not for magnesium. A time-matching algorithm was used to estimate the most accurate potassium, magnesium, and creatinine levels at exact time of AF onset. The time of AF onset was defined as the first postoperative documentation of AF/flutter in nursing flowsheets, and was converted to postoperative hours. The median AF onset time in the POAF group (50.9 hours) was used as a surrogate time point in the control group. Lab results were only used if they were within ± 12 hours from onset, or within interquartile range, and results were excluded if potassium or magnesium had been administered between lab draw and AF onset.

Demographic, perioperative, and laboratory data was compared between POAF and control groups. Means were compared using the 2 sample t test with Dunn-Sidak adjustment for multiple comparisons. The Mann-Whitney U statistic was used to compare medians and Χ2 analysis or Fisher’s exact test was used to compare categorical variables. Kaplan-Meier survival curves of freedom from POAF were compared with the log-rank test. Multivariate logistic regression analysis was used to analyze the effect of postoperative electrolyte levels on POAF occurrence while controlling for known clinical risk factors.


A total of 2,041 patients were included in the study cohort. The overall incidence of AF during postoperative course was 36.8%; median time to onset was 50.9 hours. The plasma potassium, magnesium, and creatinine curves over the first 120 postoperative hours were all significantly higher in the POAF group than in the control group.

At the time of AF onset or at the matched time point, patients with POAF had higher mean potassium and magnesium levels than control, whereas creatinine levels were not different between the 2 groups. The study cohort was then divided into quintiles based on potassium or magnesium levels at AF onset or the match time point. For potassium and magnesium, the incidence of POAF was higher in patients belonging to higher electrolyte level quintiles. However, the relation was especially notable for magnesium (72.3% incidence of POAF in the highest quintile vs 35.7% in the lowest quintile).

On univariate analysis, increased potassium, magnesium, and creatinine levels were each associated with increased risk of POAF occurrence. On multivariate logistic regression analysis, magnesium level was a strong independent predictor of AF occurrence. Other independent predictors included age, Caucasian race, preoperative β-blocker use, valve operation, and postoperative pneumonia. 

Overall, 90% of patients received potassium administration postoperatively (95% in POAF group, 87% in control group) and 22% of patients received magnesium (31% in POAF group, 17% in control). In patients who received electrolyte repletion, potassium supplementation did not affect rate of AF, whereas the rate of AF was higher in patients who received supplemental magnesium.


The link between POAF and electrolyte deficiencies has not been well established. This study suggests that patients with POAF had higher potassium and magnesium levels at time of AF onset, and supplemental magnesium may increase POAF. Therefore, current electrolyte repletion protocols may need to be further examined with a randomized trial. 

Caution should be taken in interpreting a causative relation between magnesium levels and POAF risk, as only patients with qualifying time-matched data for both potassium and magnesium could be included in logistic regression analysis, and this group of patients was at higher clinical risk of POAF based on age and prevalence of valve operation. In addition, patients who received magnesium supplementation were at higher clinical risk of POAF based on prevalence of valve operation.

Given the retrospective analysis, providers may have been biased to treat patients perceived as having higher clinical risk of POAF (based on age, valve operation, atrial ectopy) with more aggressive prophylactic electrolyte supplementation. This may have led to higher electrolyte levels in patients that were already predisposed to develop POAF. An attempt was made to control for confounders through the use of time-matched analysis, the exclusion of inaccurate laboratory results, and the use of multivariate analysis to identify clinical factors; however, the potential for confounders cannot be completely removed. The data suggests that electrolyte supplementation is ineffective in the prevention of POAF; however, it is inappropriate to conclude magnesium supplementation causes POAF. The mechanism of POAF needs to be further studied and may not be related to potassium or magnesium deficiency, as is frequently believed.

Due to the widespread prevalence of electrolyte supplementation, severe hypokalemia or hypomagnesemia was not seen, and patients in whom the electrophysiologic effects of electrolyte deficiencies may have been more apparent. Extensive cardiac literature supports the maintenance of normal to high-normal potassium and magnesium levels for prevention of ventricular arrhythmias in patients with ischemia and heart failure, and changes to current repletion practices should be approached with caution and consideration of the effect on ventricular ectopy in addition to atrial arrhythmias.1-3

  1. Horvath KA, Acker MA, Chang H, et al. Blood transfusion and infection after cardiac surgery. Ann Thorac Surg. 2013;95(6):2194–2201. doi:10.1016/j.athoracsur.2012.11.078.
  2. Quinn J, Kramer N, McDermott D. Validation of the social security death index (SSDI): an important readily-available outcomes database for researchers. West J Emerg Med. 2008;9(1):6–8.
  3. Vassar M, Holzmann M. The retrospective chart review: important methodological considerations. J Educ Eval Health Prof. 2013;10:12. doi:10.3352/jeehp.2013.10.12.