Secondary Logo

Journal Logo

Correction of Ionized Plasma Magnesium During Cardiopulmonary Bypass Reduces the Risk of Postoperative Cardiac Arrhythmia

Wilkes, Nicholas J., FRCA*,; Mallett, Susan V., FRCA*,; Peachey, Tim, FRCA*,; Di Salvo, Carmelo, MD†,; Walesby, Robin, MSc, FRCS

doi: 10.1213/00000539-200210000-00008
CARDIOVASCULAR ANESTHESIA: Research Report
Free

We conducted this randomized controlled trial to determine whether the intraoperative measurement and correction of ionized plasma magnesium can reduce the risk of cardiac arrhythmia after cardiopulmonary bypass. Eighty-five patients presenting for coronary artery bypass grafting were randomly assigned either to the magnesium-corrected group, which received magnesium sulfate on the basis of measured levels of ionized plasma magnesium (n = 43), or to the control group, in which magnesium levels were identified but not corrected (n = 42). Ionized magnesium was determined with an ion-selective electrode with minimal delay, and further samples were taken for laboratory analysis of total plasma magnesium. All patients had Holter electrocardiogram monitoring for 72 h after surgery. Total hypomagnesemia (45 patients; 53% of all patients) was more common than ionized hypomagnesemia (11 patients; 13%) before cardiopulmonary bypass. Both total and ionized magnesium levels declined further during the course of cardiopulmonary bypass in the control group. The incidence of ventricular tachycardia in the first 24 h was less frequent in the magnesium-corrected group (3 patients; 7%) than the control group (12 patients, 30%;P < 0.01). Patients in the magnesium-corrected group were more likely to display continuous sinus rhythm (Lown Grade 0) in the first 24 h (14 patients; 34%) than patients in the control group (2 patients, 5%;P < 0.001). Our results suggest that the intraoperative correction of ionized magnesium is associated with a reduction in postoperative ventricular arrhythmia in cardiac surgical patients.

Departments of *Anaesthesia and †Cardiothoracic Surgery, Royal Free Hospital, London, United Kingdom

June 11, 2002.

Address correspondence and reprint requests to Nicholas J. Wilkes, FRCA, Department of Anaesthesia, Royal Free Hospital, Pond Street, London NW3 2QG. Address e-mail to Nicholas. Wilkes@rfh.nthames.nhs.uk.

Cardiac arrhythmia after myocardial revascularization has been associated with low cardiac output, an unfavorable myocardial oxygen balance, stroke, prolonged hospital stay, and increased mortality (1,2). Electrolyte abnormalities may play an important role in the genesis of cardiac rhythm disturbances. Total plasma hypomagnesemia is common in patients undergoing cardiopulmonary bypass (CPB) (3). Despite evidence that the empirical administration of magnesium sulfate to cardiac surgical patients reduces the risk of postoperative cardiac arrhythmia (4–8), prophylactic magnesium treatment has remained controversial (9). Adverse events such as hypotension, prolongation of nondepolarizing neuromuscular block, respiratory failure, higher defibrillation energy requirements, and cardiac arrest after overdose have been related to the administration of magnesium sulfate (Table 1) (10–14).

Table 1

Table 1

The recent introduction of near-patient ion-selective electrodes (15) provides two major benefits. The technology measures the ionized, and hence, physiologically active portion of circulating plasma magnesium. The point-of-care design facilitates a flexible regimen of magnesium sulfate supplementation that is based on serial and real-time measurements of ionized magnesium, thereby allowing more accurate titration than is possible with a fixed-dosing regimen. The aim of this study was to determine whether the correction of ionized plasma magnesium, guided by measurements from an ion-selective magnesium electrode, can reduce the incidence of postoperative cardiac arrhythmia in cardiac surgical patients.

Back to Top | Article Outline

Methods

After ethics committee approval, 85 consecutive patients presenting for elective coronary artery bypass grafting (CABG) on CPB were recruited to participate in this study. Written, informed consent was obtained from each patient. Exclusion criteria were any cardiac rhythm other than sinus rhythm, on the basis of preoperative 12-lead electrocardiography (ECG); the use of drugs with antiarrhythmic properties other than β-receptor antagonists; myocardial infarction in the preceding 3 mo; poor left ventricular function, defined as an ejection fraction of <30%; impaired renal function, defined by a serum creatinine level >120 μg/dL; and endocrine or metabolic disease that may have resulted in clinically significant electrolyte imbalance.

The surgical technique was that of progressive revascularization with intermittent cross-clamping and fibrillation/defibrillation at 32°C. This is the preferred technique of revascularization used almost exclusively in the London region of England. With this technique, the bypass circuit was primed with a solution consisting of 2500 mL of Hartmann’s solution (lactated Ringer’s solution; Baxter Healthcare Ltd., Thetford, UK) and 500 mL of Gelofusin (succinylated gelatin solution; B. Braun Medical Ltd., Sheffield, UK). The oxygenator was a standard COBE laboratory flat membrane oxygenator with pump flows at 2.4 L · m−2 · min−1 (COBE Cardiovascular, Arvada, CO). Bypass was initiated between a single atrio-caval cannula and an aortic arch return. With progressive revascularization, the heart was electrically fibrillated during a short period of cross-clamping of the aorta, and the distal anastomoses were made to the coronary arteries. The aortic clamp was then released, and the heart either spontaneously defibrillated or was defibrillated, during which time the proximal anastomosis to the aorta was completed. The process was then repeated until all the vessels had been revascularized. The procedure was performed by two surgeons using identical techniques, with distal anastomoses requiring an average of 8 min of cross-clamp/fibrillation time and a total bypass time of 15 min per graft attached. During the time that the final vessel was revascularized, the patient was rewarmed after application of the aortic clamp for the final distal anastomosis, thereby returning to normothermia within 10–15 min.

The anesthetic procedure was standardized. After IV induction with fentanyl 5 μg/kg, etomidate 0.1 mg/kg, and pancuronium 0.1 mg/kg, anesthetic maintenance was based on isoflurane 0.5%–1.5% in oxygen and air before and during CPB. During rewarming, an infusion of propofol 0.1 mg · kg−1 · h−1 was initiated and continued until weaning from respiratory support in the surgical high-dependency unit. Serum potassium levels were maintained in the range of 4.5–5.0 mmol/L in all patients by repeated infusion of potassium chloride 10–20 mmol when required. Sodium heparin was administered at 300 IU/kg to achieve an activated coagulation time of more than 450 s. Protamine 3 mg/kg was given for reversal of systemic heparinization. Morphine was administered at 0–150 μg · kg−1 · h−1 to provide appropriate levels of analgesia in the first 24 postoperative hours. Patients who received regular β-receptor antagonist medication were given their morning dose on the day of the operation. Oral medication was withheld for 24 h after surgery. β-Receptor antagonists were recommenced on the second postoperative day.

Ionized and total plasma magnesium were measured at 6 specified times (before surgery, twice during CPB cold, during CPB warm, during surgery after CPB, and within 1 h after surgery). Blood samples were obtained from an indwelling radial artery catheter. For the determination of ionized magnesium, samples were collected in a heparinized syringe before and after CPB and in an unheparinized syringe during CPB and were analyzed immediately by using an ion-selective electrode (Electrolyte 8 Analyzer; Nova Biomed, Waltham, MA). This technology uses a neutral carrier-based membrane that has an ionophore selective for the size of the magnesium ion (16). Blood samples for analysis of total plasma magnesium were stored in Sarstedt Monovette® serum gel blood collection tubes and analyzed using a colorimetric method with a Hitachi 747 Analyzer (17). The reference ranges were 0.45–0.60 mmol/L for ionized plasma magnesium and 0.70–1.0 mmol/L for total plasma magnesium.

All patients were monitored for 72 h with a two-channel Holter monitor providing dual-lead electrographic recordings (leads CM2 and CM5; Tracker 2; Reynolds Medical, Hertfort, UK) (18). An independent cardiology technician who was blinded to the patient’s group allocation and treatment performed the analysis of the recordings. Artifact was excluded by visual analysis of irregular rhythm. Arrhythmia was classified as supraventricular or ventricular in origin. Atrial fibrillation (AF), supraventricular tachycardia, ventricular tachycardia (VT), and ventricular fibrillation were considered serious arrhythmia. Sustained arrhythmia was diagnosed when treatment was required to reestablish sinus rhythm. Nonsustained arrhythmia was defined as five or more irregular beats on Holter monitoring. The Lown classification was applied to sinus rhythm with ventricular premature beats (VPB) (Table 2).

Table 2

Table 2

Patients were allocated to one of two groups by random numbers generated from random-number tables. The magnesium-corrected group received supplements of magnesium sulfate on the basis of mea-sured ionized magnesium levels aiming to prevent hypomagnesemia and achieving ionized magnesium levels in the upper normal range. In a pilot study we had determined that 10–15 mmol of magnesium sulfate was needed during CPB to achieve normal or supernormal levels of ionized magnesium that would return to the reference range in the postoperative period (19). Intraoperative levels of ionized magnesium <0.50 mmol/L were treated with magnesium sulfate 10 mmol suspended in 50 mL of sodium chloride 0.9% solution, administered over 20 min. This intervention was repeated as required. In the control group, ionized magnesium levels were measured but not treated. To administer magnesium sulfate on the basis of ionized magnesium plasma levels and repeat this intervention as required, one nominated investigator remained unblinded to treatment group. All other clinicians involved in the care of patients and technicians concerned with the analysis of Holter tapes were rigorously blinded throughout the study period.

On the basis of previous publications investigating postoperative cardiac arrhythmia after CPB, we anticipated a frequency of arrhythmia of 40% in untreated patients and assumed a 50% reduction after treatment with magnesium sulfate (4,5). We calculated that we would need a total sample size of 80 to show that the difference was statistically significant at the 5% level with 80% power. We used paired and unpaired Student’s t-tests for normally distributed interval data and χ2 tests for nominal data.

Back to Top | Article Outline

Results

Demographic data were similar between the study groups (Table 3). No patients with poor left ventricular function (left ventricular ejection fraction <30%) had been included in the trial. Two patients died of myocardial infarction in the immediate postoperative period (one in each study group). Another two patients underwent mediastinal reexploration for bleeding complications (one in each group). All of these patients were excluded from further analysis of postoperative cardiac rhythm.

Table 3

Table 3

The mean preoperative ionized magnesium levels in all patients were 0.50 ± 0.05 mmol/L (Fig. 1). Eleven patients (13% of all patients) displayed preoperative ionized hypomagnesemia, and four patients (5%) demonstrated ionized hypermagnesemia. The mean preoperative total magnesium levels were 0.69 ± 0.09 mmol/L (Fig. 2). Forty-five patients (53%) presented with total magnesium levels below the reference range when tested before surgery, whereas none had high total magnesium levels.

Figure 1

Figure 1

Figure 2

Figure 2

A significant decline in both ionized (to 0.43 ± 0.05 mmol/L;P < 0.001) and total magnesium (to 0.56 ± 0.07 mmol/L;P < 0.001) during and after CPB ensued in patients in the control group. This was followed by a mild, but statistically significant, recovery (to 0.45 ± 0.04 and 0.59 ± 0.11 mmol/L, respectively; both P < 0.02) immediately after surgery (Figs. 1 and 2). Thirty of 40 patients (75%) in the control group had developed ionized hypomagnesemia immediately after CPB, and all of the patients in this group showed total hypomagnesemia at this measurement point.

All patients in the magnesium-corrected group received magnesium sulfate supplements (mean, 13.4 ± 0.9 mmol; range, 10–30 mmol). In this group, mean ionized magnesium levels were 0.59 ± 0.08 mmol/L, and total magnesium levels were 0.94 ± 0.16 mmol/L immediately after surgery.

Twelve of 40 patients (30%) in the control group demonstrated episodes of VT in the first 24 postoperative hours (Table 2). This finding compares with 3 of 41 patients (7%) in the magnesium-corrected group. The difference was statistically significant (P < 0.01) and indicated a reduction in the risk of postoperative arrhythmia by 76% in the magnesium-corrected group.

Fourteen of 41 patients (34%) in the magnesium-corrected group displayed uninterrupted sinus rhythm after surgery (Lown Grade 0) compared with 2 of 40 patients (5%) in the control group (P < 0.001). Patients with uninterrupted sinus rhythm had significantly higher postoperative ionized and total magnesium levels than patients who developed serious postoperative cardiac arrhythmia or demonstrated VPB (Lown Grades 1–4: ionized magnesium, 0.57 ± 0.06 mmol/L versus 0.51 ± 0.05 mmol/L, P < 0.01; total magnesium, 0.93 ± 0.13 mmol/L versus 0.75 ± 0.12 mmol/L, P < 0.01).

Back to Top | Article Outline

Discussion

The results of our study demonstrate that the intraoperative correction of ionized plasma magnesium can reduce the risk of cardiac arrhythmia after CPB. Patients who received magnesium sulfate for the treatment of ionized hypomagnesemia had a significantly less frequent incidence of VT and a more frequent incidence of uninterrupted sinus rhythm in the first 24 hours than patients who received no treatment. Regardless of treatment group, patients in continuous sinus rhythm without VPB had significantly higher total and ionized plasma magnesium levels than all other patients.

Magnesium is an important determinant of the resting membrane potential of cardiac cell membranes. It regulates the sodium-potassium-adenosine triphosphatase pump, thus affecting the intracellular/extracellular potassium ratio and the membrane resting potential of myocardial cells (20). Extracellular magnesium deficiency results in the loss of cellular potassium and gain of cellular sodium, leading to an increase in myocardial excitability (15). Magnesium inhibits the influx of calcium through sarcolemmal channels and modulates cyclic adenosine monophosphate, thus blocking the slow inward calcium current, whereas hypomagnesemia results in the increase of intracellular calcium.

Magnesium is a constituent of cardioplegia solution, and as such it has cardioprotective properties (21). During periods of ischemia, levels of adenosine triphosphate decrease, leading to a passive loss of intracellular magnesium. This loss is reduced when extracellular magnesium levels are increased, facilitating postischemic high-energy phosphate production. Although no cardioplegia solution was used in our study, it is possible that the antiarrhythmogenic effect of magnesium supplementation relies on similar cellular mechanisms.

In an experimental model, Du Toit and Opie (22) showed that the presence of magnesium during or soon after reperfusion after coronary occlusion prevents cellular injury related to massive calcium influx. Such evidence, and further suggestions that cardiac arrhythmia is most common within the first 24 hours after cardiac surgery (6), led us to investigate a regimen that consisted of the intraoperative supplementation of magnesium sulfate rather than a postoperative magnesium infusion. Although we initiated Holter ECG monitoring for 72 hours after surgery, we observed a difference in the incidence of cardiac arrhythmia between patient groups only in the first 24 postoperative hours. Not all patients’ data were analyzable for the entire 72-hour study period, though, and this may have contributed to the absence of intergroup differences on Days 2 and 3.

Our findings correspond with previous studies that have demonstrated an antiarrhythmogenic effect when magnesium was administered empirically, rather than in a targeted fashion. Harris et al. (4) found an incidence of arrhythmia of 22% after 16 mmol of magnesium chloride, compared with 63% in control patients. The authors measured plasma and urine magnesium. England et al. (5) discovered that the administration of magnesium chloride 2 g after termination of CPB resulted in a reduction in the incidence of ventricular arrhythmia from 34% to 16%. This group first used ultrafilterable magnesium determinations as a reflection of ionized and chelated magnesium, thus excluding the protein-bound fraction. Casthely et al. (6) administered four different regimens of magnesium sulfate during CPB and found that the highest magnesium regimen was associated with the least frequent incidence of postoperative cardiac arrhythmia. In these studies, magnesium administration prevented ventricular, but not supraventricular, rhythm abnormalities. Other authors found a greater effect on supraventricular arrhythmia: Wistbacka et al. (7) demonstrated an incidence of AF in 24% of patients who received a high magnesium infusion regimen compared with 45% of patients receiving a low-magnesium regimen. Fanning et al. (8) demonstrated a decrease in the frequency of postoperative AF in magnesium-treated patients compared with placebo but found no difference in the incidence of ventricular arrhythmia. Balser (23) concluded that small-dose postoperative magnesium supplementation reduced the incidence of ventricular arrhythmia, but larger doses of magnesium were needed to prevent AF.

Postoperative β-receptor antagonist withdrawal has been associated with an increased incidence of postoperative cardiac arrhythmia (24). Preoperative use of β-receptor antagonists was slightly more frequent in the magnesium-corrected group of our study than the control group, but the difference was not large enough to explain a difference in postoperative cardiac arrhythmia between groups.

None of the studies that were available to us have previously demonstrated a close relationship between postoperative sinus rhythm without VPB (Lown Grade 0) and higher levels of total and ionized plasma magnesium. This finding suggests that the targeted magnesium administration to cardiac surgical patients not only prevents potentially serious complications such as VT, but also reduces the frequency of VPB, considered less indicative of substantial myocardial pathology (25).

Despite the evidence that the administration of magnesium sulfate can reduce the incidence of postoperative cardiac arrhythmia, its use has remained controversial: Parikka et al. (26) demonstrated no reduction in the incidence of AF in magnesium-treated patients and concluded that high magnesium levels may even trigger AF. Hecker et al. (14) found that more direct current shocks and higher energy levels were required for defibrillation after CPB in magnesium-treated patients, indicating that high intraoperative plasma levels of magnesium may have adverse effects on the heart. In a review article of the physiology and pharmacology of magnesium and its role in anesthetic practice, Fawcett et al. (9) suggested that the main role for magnesium in cardiac surgery consisted of myocardial protection and not the prevention of postoperative arrhythmia.

In our study, a large proportion of patients presented with preoperative total hypomagnesemia (53% of all patients). A much smaller number demonstrated preoperative ionized hypomagnesemia (13%). England et al. (5) found that 16% of their patients had preoperative total hypomagnesemia and that 11% had preoperative deficiency of ultrafilterable (ionized and chelated) magnesium. Dietary magnesium deficiency has increased steadily in recent decades and is common in an urban population from which our study patients were derived (15). It has been suggested that hypomagnesemia is a consequence of pharmacological factors in patients with coronary artery disease, but it may also be a predisposing factor for the development of ischemic changes (15). Diuretic use is associated with reduced tubular reabsorption of magnesium and a total body magnesium deficiency, but only 16% of our study population had received regular diuretic medication. The development of total plasma hypomagnesemia during cardiac or hepatic surgery may be a consequence of hemodilution, increased urinary losses, an intracellular shift, high plasma citrate levels, or catecholamine-induced lipolysis with chelation of magnesium by free fatty acids (27–29).

Some methodological aspects of our study warrant comment. In all of our patients, the surgical technique incorporated intermittent cross-clamping and fibrillation/defibrillation at 32°C. Our results, therefore, should not be extrapolated to a method that uses cardioplegia. One investigator had to remain unblinded to patients’ group allocation to administer magnesium sulfate in response to mea-surements of ionized magnesium. For the purpose of excluding treatment or interpretation bias, none of the clinicians involved in patient care or the technicians concerned with measurements of magnesium levels or evaluation of Holter ECGs had knowledge of the treatment group. We deliberately investigated a regimen of intraoperative magnesium supplementation with the intention of providing high physiological levels of magnesium during reperfusion. Studies of postoperative magnesium supplementation may yield different results. Finally, we did not design this study to test for true outcome variables, such as survival, time in the critical care unit, or hospital resource utilization. Our primary outcome was the incidence of postoperative cardiac arrhythmia. This may be considered a surrogate outcome variable, because it has been directly related to hospital stay and mortality by other investigators (1,2).

In conclusion, the measurement of ionized magnesium and the correction of ionized hypomagnesemia with adjusted aliquots of magnesium sulfate contribute to the prevention of postoperative cardiac arrhythmia. Both ionized and total plasma hypomagnesemia are common in patients presenting for CABG. If not supplemented, both these values decline further during CPB. We found ion-selective magnesium electrodes to be convenient near-patient instruments that facilitated immediate determination and targeted correction of ionized plasma magnesium levels.

The authors wish to thank the Department of Cardiology at the Royal Free Hospital for processing and evaluating all Holter ECG tapes.

Back to Top | Article Outline

References

1. Atlee JL III. Anaesthesia and cardiac electrophysiology. Eur J Anaesthesiol 1985; 2: 215–56.
2. Forrest JB, Rehder K, Cahalan MK, Goldsmith CH. Multicenter study of general anesthesia. III. Predictors of severe perioperative adverse outcomes. Anesthesiology 1992; 76: 3–15.
3. Aglio LS, Stanford GG, Maddi R, et al. Hypomagnesaemia is common following cardiac surgery. J Cardiothorac Vasc Anesth 1991; 5: 201–8.
4. Harris MN, Crowther A, Jupp RA, Aps C. Magnesium and coronary revascularization. Br J Anaesth 1988; 60: 779–83.
5. England MR, Gordon G, Salem M, Chernow B. Magnesium administration and dysrhythmias after cardiac surgery: a placebo-controlled, double-blind, randomized trial. JAMA 1992; 268: 2395–402.
6. Casthely PA, Yoganathan T, Komer C, Kelly M. Magnesium and arrhythmias after coronary artery bypass surgery. J Cardiothorac Vasc Anesth 1994; 8: 188–91.
7. Wistbacka JOM, Koistinen J, Karlqvist KEV, et al. Magnesium substitution in elective coronary artery surgery: a double-blind clinical study. J Cardiothorac Vasc Anesth 1995; 9: 140–6.
8. Fanning WJ, Thomas CS Jr, Roach A, et al. Prophylaxis of atrial fibrillation with magnesium sulphate after coronary artery bypass grafting. Ann Thorac Surg 1991; 52: 529–33.
9. Fawcett WJ, Haxby EJ, Male DA. Magnesium: physiology and pharmacology. Br J Anaesth 1999; 83: 302–20.
10. Gambling DR, Birmingham CL, Jenkins LC. Magnesium and the anaesthetist. Can J Anaesth 1988; 35: 644–54.
11. Nakaigawa Y, Akazawa S, Shimizu R, et al. Effects of magnesium sulphate on the cardiovascular system, coronary circulation and myocardial metabolism in anaesthetized dogs. Br J Anaesth 1997; 79: 363–8.
12. Feldman S, Karalliedde L. Drug interactions with neuromuscular blockers. Drug Saf 1996; 15: 261–73.
13. Vissers RJ, Purssel R. Iatrogenic magnesium overdose: two case reports. J Emerg Med 1996; 14: 187–91.
14. Hecker BR, Lake CL, Kron IL, et al. Influence of magnesium ion on human ventricular defibrillation after aortocoronary bypass surgery. Am J Cardiol 1985; 55: 61–4.
15. Altura BM, Altura BT. Role of magnesium in patho-physiological processes and the clinical utility of magnesium ion selective electrodes. Scand J Clin Lab Invest Suppl 1996; 56: 211–34.
16. Altura BT, Burack JL, Cracco RQ, et al. Clinical studies with the NOVA ISE for IMg2+. Scand J Clin Lab Invest Suppl 1994; 217: 53–67.
17. Elin RJ. Determination of serum magnesium concentration by clinical laboratories. Magnes Trace Elem 1991; 10: 60–6.
18. O’Kelly B, Browner WS, Massie B, et al. Ventricular arrhythmias in patients undergoing noncardiac surgery: the Study of Perioperative Ischemia Research Group. JAMA 1992; 268: 217–21.
19. Wilkes NJ, Peachey T, Mallett SV. Measurement and correction of ionized plasma magnesium during cardiopulmonary bypass [abstract]. Br J Anaesth 1998; 80 (Suppl 1): A289.
20. Elin RJ. Magnesium: the fifth but forgotten electrolyte. Am J Clin Pathol 1994; 102: 616–22.
21. Hearse DJ, Stewart DA, Braimbridge MV. Myocardial protection during ischaemic arrest. J Thorac Cardiovasc Surg 1978; 75: 877–85.
22. Du Toit EF, Opie LH. Modulation of severity of reperfusion stunning in the isolated rat heart by agents altering calcium flux at onset of reperfusion. Circ Res 1992; 70: 960–7.
23. Balser JR. Pro: all patients should receive pharmacologic prophylaxis for atrial fibrillation after cardiac surgery. J Cardiothorac Vasc Anesth 1999; 13: 98–100.
24. Bert AA, Reinert SE, Singh AK. A beta-blocker, not magnesium, is effective prophylaxis for atrial tachyarrhythmias after coronary artery bypass graft surgery. J Cardiothorac Vasc Anesth 2001; 15: 204–9.
25. Knopes KD, Hecker BR. Magnesium therapy in cardiac surgery: con—magnesium is not a valuable therapy in the cardiac surgical patient. J Cardiothorac Vasc Anesth 1991; 5: 522–4.
26. Parikka H, Toivonen L, Pellinen T, et al. The influence of intravenous magnesium sulphate on the occurrence of atrial fibrillation after coronary artery by-pass operation. Eur Heart J 1993; 14: 251–8.
27. Scheinman MM, Sullivan RW, Hyatt KH. Magnesium metabolism in patients undergoing cardiopulmonary bypass. Circulation 1969; 34: 235–41.
28. Ranasinghe DN, Mallett SV. Hypomagnesaemia, cardiac arrhythmias and orthotopic liver transplantation. Anaesthesia 1994; 49: 403–5.
29. Külpmann WR, Rademacher E, Bornscheuer A. Ionized magnesium concentration during liver transplantation, resection of the liver and cardiac surgery. Scand J Clin Lab Invest Suppl 1996; 224: 235–43.
© 2002 International Anesthesia Research Society