0
Clinical Investigations: CARDIOLOGY |

Admission Serum Potassium in Patients With Acute Myocardial Infarction*: Its Correlates and Value as a Determinant of In-Hospital Outcome FREE TO VIEW

John E. Madias, MD; Bimal Shah, MD; Gopal Chintalapally, MD; Gopal Chalavarya, MD; Nicolaos E. Madias, MD
Author and Funding Information

*From the Mount Sinai School of Medicine of New York University, and the Division of Cardiology (Drs. J. Madias, Shah, Chintalapally, and Chalavarya), Elmhurst Hospital Center, Elmhurst, NY; and the Tufts University School of Medicine, and the Division of Nephrology (Dr. N. Madias), New England Medical Center, Boston, MA.

Correspondence to: John E. Madias, MD, Division of Cardiology, Elmhurst Hospital Center, 79–01 Broadway, Elmhurst, NY 11373, e-mail: jmad@pop.nychhc.org



Chest. 2000;118(4):904-913. doi:10.1378/chest.118.4.904
Text Size: A A A
Published online

Study objectives: Although controversial, hypokalemia (LK) in patients with acute myocardial infarction (MI) is thought to predict increased in-hospital morbidity, particularly cardiac arrhythmias, and mortality. Also, the mechanism of low serum potassium in the setting of MI has not been delineated. We evaluated the frequency, attributes, and outcome, and speculated on the mechanism of LK in patients with MI.

Design: This was a prospective cross-sectional study of 517 consecutive patients with MI admitted to the coronary care unit (CCU). Serum potassium was measured in the emergency department and repeatedly thereafter throughout hospitalization, and was used in the analysis, along with a large array of clinical and laboratory variables.

Results: The patients were allocated to a LK and a normokalemic (NK) cohort, based on the emergency department serum potassium measurement. The 41 patients with LK (3.16 ± 0.24 mEq/L; 7.9% of total) were comparable on admission in their baseline assessment to the 476 patients with normal serum potassium (4.28 ± 0.56 mEq/L), except for lower emergency department magnesium (1.48 ± 0.15 mg/dL vs 1.96 ± 0.26 mg/dL; p = 0.0005) and earlier presentation after onset of symptoms (3.0 ± 4.1 h vs 4.4 ± 6.2 h; p = 0.05). There was a poor correlation between serum potassium and magnesium on admission (r = 0.14). Peak creatine kinase (CK) and myocardial isomer of CK were higher in the LK patients (3,870 ± 3,840 IU/L vs 2,359 ± 2,653 IU/L [p = 0.018] and 358 ± 312 IU/L vs 228 ± 258 IU/L [p = 0.013], respectively). Management of the two cohorts was the same, except for a higher rate of use of magnesium (14.6% vs 4.6%; p = 0.007), serum potassium supplements (90.2% vs 43.1%; p = 0.000005), and antiarrhythmic drugs (78.0% vs 50.4%; p = 0.0007) in the LK patients. No difference was detected between the LK and NK patients in total mortality (24.4% vs 18.3%; p = 0.34), cardiac mortality (17.1% vs 15.3%; p = 0.52), atrial fibrillation (14.6% vs 13.9%; p = 0.89), and ventricular tachycardia (22.0% vs 16.0%; p = 0.32), but ventricular fibrillation (VF) occurred more often (24.4% vs 13.0%; p = 0.04) in the LK patients. However, proportions of VF occurring in the emergency department, CCU, or wards in the two cohorts were not different, but they were higher during the time interval prior to emergency department admission in LK patients (17.1% vs 2.1%; p = 0.00001).

Conclusions: LK is seen in approximately 8% of patients with MI in the emergency department; LK is associated with low emergency department magnesium, and low serum potassium levels in the CCU and throughout hospitalization. LK has no relationship to preadmission use of diuretics, it is associated with early presentation to the emergency department, and it is not a predictor of increased morbidity or mortality.

The role of hypokalemia (LK) in cardiovascular disease in general, or in myocardial ischemia and myocardial infarction (MI) in particular, has been under investigation for a long time.19 Drawing from the vast literature, LK is found in 9 to 25% of patients with MI,1013 and it is portrayed as a determinant of excessive morbidity, particularly malignant ventricular arrhythmia (VA), or mortality.1011,1423 Also, results on the effect of treatment with diuretics in predisposing to LK on admission of patients with MI have been variable.8,11,13,2326 Controversy about LK in MI exists in a context of persisting disagreement about a potentially harmful effect of LK, per se.14,2733

Close examination of the literature on LK in patients with MI reveals the complexity of the matter, with its inherent ample chance for methodologic and clinical pitfalls in the design and implementation of studies. Thus, existing works have been mostly retrospective,7,14,16,23 although some looked at this topic prospectively.19,21,34 The studies have not always provided data on the time of serum potassium measurement, which was either not specified,7,27 vaguely noted,14 or occurred within a range of many hours7 or days23,35 after admission, with few exceptions.19,34 The temporal relation of serum potassium measurements with the time of onset of illness and emergency department admission, or the occurrence of complications, has often been missing.7,27 The preadmission and hospital study parameters have often been few, have been limited only to a very sketchy characterization of the patients, and have focused merely on ventricular tachycardia (VT) and ventricular fibrillation (VF) as the sole outcome end points.7,11,24 Details on therapy have often been missing entirely78; particularly, no information has been furnished on serum potassium supplementation, or use of β-blockers and antiarrhythmic drugs.78,26 Also, as a rule, the definition of employed variables has not been rigorous.20 Finally, even the threshold for the diagnosis of LK has varied widely, employing as cutoff points serum potassium levels < 3.5, < 3.6, or < 4.0 mEq/L.1011,13,16,23,36 This has often rendered comparison between studies meaningless.19 While some of these issues have been adequately dealt with in some publications,19,34 there is no work addressing all of the issues in a single study.

To reassess LK in MI, we embarked on a study of patients admitted during a period of 3.5 years. Our objectives were to evaluate the following: (1) the incidence of LK in patients with MI on presentation to the emergency department; (2) the relationship of emergency department-measured serum potassium with prior use of diuretics; (3) other variables that could be associated with emergency department measurements in LK; (4) the changes of serum potassium values after emergency department measurement in the LK and normokalemic (NK) cohorts; (5) the rate of complications of LK and NK patients; (6) the timing, location, and the circumstances under which VT and VF occurred in the study patients; and (7) determination of a plausible mechanism of LK.

Patient Selection

Consecutive patients with MI who were admitted to the coronary care unit (CCU) between January 1, 1986, and June 30, 1989, were prospectively evaluated; the database is the same as used in another published work.37The diagnosis of MI was substantiated by suggestive symptoms, sequential changes in serial ECGs,3839 peak creatine kinase (CK) greater than two times the upper reference value of our laboratory, and peak myocardial isomer of CK (CK-MB) > 6%. CK enzymes were measured thrice daily for 2 days, every day for 5 days, and as clinically indicated thereafter.

Collection of Admission Data

Information gathered by ambulance personnel was sought, reviewed, and included in the database. Data obtained in the emergency department, CCU, and wards were entered in custom-designed forms and then to a computer database by our research fellows. Variables included, among others, routine demographics and risk factors, and prior use of diuretics, serum potassium supplements, digitalis, and cardioselective or noncardioselective β-blockers. Inquiry was made about the time interval between the onset of symptoms and arrival in the emergency department. Evaluation in the emergency department and CCU also comprised admission Killip class,40 and assessment for pulmonary congestion or cardiomegaly on the chest radiogram. Occurrence and management of VT or VF in the field, ambulance, or emergency department were reviewed and included in the database. One of the authors (JEM) interpreted all ECG rhythm strips.

Serum potassium was measured upon arrival of the patients in the emergency department, using a potensiometric method with an ion-selective electrode41; the calmagite spectrophotometric method was used to measure magnesium.42 LK was defined as serum potassium < 3.5 mEq/L. Further measurements of serum potassium were obtained on transfer to the CCU, and repeatedly thereafter at the discretion of attending physicians for monitoring purposes, or prophylactically for patients treated with diuretics, or to assess the response to serum potassium replenishment. Data analysis comprised the emergency department- and CCU-measured serum potassium, and a mean of all serum potassium values obtained during hospitalization. The care of patients was not influenced by the study design or the investigators, but followed the usual CCU procedures. Serum potassium supplementation for LK often started even prior to transfer to the CCU, while administration of magnesium salts for hypomagnesemia always commenced in the CCU. Because at the time of the study we were operating as a collaborative site for an intracoronary thrombolysis protocol,43 and IV thrombolysis was not routinely employed, only a fraction of the eligible study patients received such therapy.

Collection of Hospitalization Data

Information was secured on peak CK and CK-MB levels (taken as rough indicators of the magnitude of MI),44 ECG site of MI, ejection fraction (using radionuclide ventriculography), and number of coronary vessels with ≥ 70% diameter stenosis at coronary arteriography, performed prior to discharge. Data on the use of IV heparin, nitroglycerin, calcium-channel inhibitors, β-blockers, thrombolytics, antiarrhythmic drugs, oral organic nitrates, digitalis, diuretics, serum potassium (as IV and oral KCl), and magnesium (as IV MgSO4) supplements were collected as they became available. The time of transfer of a patient to the CCU from the emergency department depended on the patient’s clinical condition or CCU bed availability.

Patient outcome variables included total and cardiac mortality, peak Killip class reached during hospitalization, extension of MI, post-MI angina, post-MI angina associated with transient ischemic ECG changes, atrial fibrillation, atrial flutter, frequent premature ventricular contractions (PVCs), VT, VF, and use of cardioversion for sustained VT. Details were sought regarding the site of occurrence of VT or VF (pre-emergency department, in the emergency department, in the CCU, or in the ward). In addition to total VT and VF, we implemented a subgrouping of VT and VF episodes to primary (associated with Killip I), secondary (Killip class II-IV), and spontaneous (as differentiated from the ones emerging during cardiac catheterization, insertion of temporary pacing leads or Swan-Ganz catheters, or injections of contrast medium). Radiographic diagnosis of congestion or cardiomegaly was based on the official reports of the radiology department. All MIs of Q-wave type that were not diagnosed as anterior or inferior/inferolateral were categorized as “other.” This designation included patients with left bundle-branch block, transient repolarization changes often associated with non-Q-wave MI,38 and unaltered ECGs, by comparison with previous tracings. Cardiac mortality (in contrast to total) denoted death from pump failure, cardiac standstill, or VT and VF. “Frequent PVCs” were diagnosed when such VA led to a comment by a physician in the patient’s record, or when it resulted in initiation of antiarrhythmic therapy. VA occurring during thrombolysis was categorized as spontaneous, although it could have been triggered by reperfusion. Occurrences of idioventricular rhythm (slow VT) were ignored. Patients were followed until death or discharge from the hospital.

Statistical Analysis

All data were considered in the analysis. Cohort data are reported as mean ± SD, or percentage occurrences of a variable. Cohort means of ratio scale variables were compared by two-tailed Student’s t test for paired and unpaired data, as appropriate. The reported p values were based on the result of the Levine test to decide whether the pooled or the separate variance estimates should be used. Nominal and ordinal scale variables were analyzed by χ2 or Fisher’s Exact Test, as appropriate. The rate of VT and VF at pre-emergency department, emergency department, CCU, or wards were analyzed by 2 × 4 contingency tables and the χ2 test. Correlation between variables employed Pearson’s correlation coefficient.45Relations between serum potassium and magnesium values at different time settings were analyzed separately for all patients, or the LK and the NK cohorts. Statistical significance was accepted when p = 0.05. The SPSS/PC + 4.0.1 statistical package (SPSS; Chicago, IL) was used in the analysis.4647

The database contained 517 patients; their emergency department serum potassium level was 4.19 ± 0.62 mEq/L (range, 2.5 to 7.4 mEq/L). Comparison of emergency department serum potassium levels in 393 of the 517 patients who had repeat serum potassium measurements in the CCU with the patients who did not revealed no significant difference (4.14 ± 0.63 mEq/L vs 4.09 ± 0.52 mEq/L; p = 0.1). Also, emergency department serum potassium levels for 329 patients who had a mean serum potassium calculated did not reveal a significant change between these two values (4.14 ± 0.64 mEq/L vs 4.09± 0.35 mEq/L; p = 0.36). In the same 329 patients, serum potassium levels in the CCU and mean serum potassium were not different (4.08 ± 0.53 mEq/L vs 4.09 ± 0.35 mEq/L; p = 0.53). Correlations between emergency department serum potassium and CCU serum potassium (r = 0.43; p = 0.0005), and emergency department serum potassium and mean serum potassium (r = 0.31; p = 0.0005) were fair, while correlation between CCU serum potassium and mean serum potassium (r = 0.66; p = 0.0005) was good. Correlation of the emergency department serum potassium and magnesium levels in all 517 patients (r = 0.14; p = 0.002) was poor.

Although not by study design, a number of patients had measurements of serum bicarbonate and arterial blood gases in the emergency department. There was no difference between LK and NK patients in reference to these measurements (Table 1 ).

Of the 517 patients with LK, 41 patients (7.9% of total) had LK in the emergency department, with serum potassium levels averaging 3.16 ± 0.24 mEq/L; the remaining 476 patients had NK, with emergency department serum potassium averaging 4.28 ± 0.56 mEq/L (p = 0.005). LK was detected in the emergency department in 8.8% of men and 5.9% of women (p = 0.26). Thirty-eight of the 41 LK patients and 355 of the 476 NK patients had a repeat serum potassium measurement in the CCU. Also, 37 of the 41 LK patients and 292 of the 476 NK patients had a mean serum potassium calculated; these data are shown in Table 1. Magnesium values in the emergency department were 1.48 ± 0.15 mg/dL and 1.96 ± 0.26 mg/dL in the LK and NK groups, respectively (p = 0.0005).

Comparisons of the emergency department serum potassium values in 38 of the 41 LK patients with serum potassium measured in the CCU revealed significant increase in the latter (3.16 ± 0.25 mEq/L vs 3.78 ± 0.48 mEq/L; p = 0.0005). Also, a comparison of emergency department serum potassium and mean serum potassium levels in 37 patients showed a significant increase in the latter (3.16 ± 0.25 mEq/L vs 3.86 ± 0.35 mEq/L; p = 0.0005); in the same group, there was a significant but modest change between the CCU serum potassium and mean serum potassium values (3.78 ± 0.46 mEq/L vs 3.86 ± 0.35 mEq/L; p = 0.037). Correlation of the emergency department measurement of serum potassium and magnesium in the LK patients (r = 0.38; p = 0.015) was fair.

Patients With NK

In contrast, 355 of the 476 NK patients revealed a significant drop in serum potassium values between the emergency department and the CCU (4.25 ± 0.56 mEq/L vs 4.13 ± 0.51 mEq/L; p = 0.0005), and in 292 patients between emergency department serum potassium and mean serum potassium (4.25 ± 0.56 mEq/L vs 4.12 ± 0.35; p = 0.0005). No changes were detected in these 292 patients between the CCU serum potassium and the mean serum potassium levels (4.12 ± 0.52 mEq/L vs 4.12 ± 0.35 mEq/L; p = 0.953). Correlation of the emergency department serum potassium and magnesium values in the NK patients (r = 0.09; p = 0.051) was poor.

Attributes of Patients With LK and NK

Demographic, clinical, and laboratory characteristics of the patients are shown in Table 1. Only a few significant differences were noted between the two cohorts. Thus, a lower CCU serum potassium, mean serum potassium, and emergency department magnesium measurements were found in the LK group; also, these patients had a shorter time interval (mean, 1.4 h) between onset of symptoms and presentation to the emergency department, and higher peak values of CK and CK-MB.

Clinical Management

As shown in Table 2 , in-hospital management was similar in the two cohorts, except for more frequent use of serum potassium and magnesium supplements, antiarrhythmic drugs, and lower employment of heparin and IV nitroglycerin in the LK patients. Thrombolysis was given to 88 of the 336 eligible patients (26.2%) with a Q-wave MI; such therapy was not implemented in the 32 patients with left bundle-branch block, as was the standard at the time of the study.

Clinical Course

Table 3 details the clinical outcome of the two cohorts. Total and cardiac mortality were not significantly different, and the frequency of supraventricular arrhythmias and VT were similar in the two groups. Also, the proportions of total VT occurring at pre-emergency department, emergency department, CCU, and wards were the same in the two cohorts. In contrast, total VF was more frequent in the LK patients than in the NK patients; this difference was due to the increased rate of VF occurring at pre-emergency department, while the rates of VF in the other three locations were the same for the LK and NK cohorts. Importantly, in 9 of 10 patients who had VF and were found to have LK after resuscitation (7 at pre-emergency department and 2 in the emergency department), only 3 patients took diuretics prior to admission. The remaining patient with VF in the CCU had not used diuretics.

The following conclusions derive from this study: (1) LK in the emergency department is found in approximately 8% of patients with MI; (2) in comparison with NK, LK is not associated with prior use of diuretics; (3) LK is seen with an earlier presentation to the emergency department; (4) LK is linked to larger MIs; and (5) LK is not associated with a higher mortality, or other complications, except for an increased rate of VF. However, since 9 of the 10 patients with VF and LK had this arrhythmia at pre-emergency department or emergency department settings, and only 3 of them were taking diuretics, LK may reflect a postresuscitation epiphenomenon and/or may be related to an early presentation in the emergency department, rather than be the cause of VF.

Mechanism of LK in MI

In our study, the association of LK after resuscitation from VF with early presentation to the emergency department, large MIs, and no link to diuretic therapy suggests a “stress”-induced intracellular serum potassium shift as the explanation for LK.4850 Its mechanism has been traced to a β2-adrenergic stimulation in normal volunteers and borderline hypertensive patients.54 Stress leads to sympathetic overstimulation of the adrenal medulla, with outpouring of catecholamines, predominantly epinephrine, functioning as hormones; in contrast, such mediators released at terminal fibers of sympathetic nerves act locally.5556 The contribution of the possible variation in the use and amount of epinephrine (injected by the ambulance personnel in the management of VF in the 7 patients with LK and 10 patients with NK) to the eventual serum potassium level could not be examined, since this was not part of the original protocol design, and the patients records are no longer available to the investigators. The β2-adrenoceptor stimulation is linked to membrane-bound Na+/K+ adenosine triphosphatase, which in turn mediates an active serum potassium influx to the skeletal muscles.,54 Such serum potassium shifts are expected to be transient and reversible, to develop quickly, and to depend on an equally transient adrenergic surge. LK in both animals and human volunteers developed within 1 h of catecholamine infusions.51,53,57 Nonselective andβ 2-selective blockers are reputed to protect patients exposed to high natural or artificial catecholamime states from developing LK, while β1-selective blockers do not.,5152 Some workers have reported that β-blockers do not completely prevent LK.25,34 Although we interpreted the higher CK and CK-MB levels in the LK subgroup as indicative of larger MIs, the similarity in the clinical outcome in the two subgroups did not corroborate this assertion (Table 3).

Epinephrine in humans accentuates the LK resulting from use of diuretics.58LK has been noted as a side effect ofβ 2-adrenergic agonists used as bronchodilators.59 Such agonists have been implemented in the management of hyperkalemia.6061 In both these situations, intracellular serum potassium shifts are thought to be at work.

A nonspecific body response to stress is suggested by LK documented in patients with the following: MI, chest pain without MI,35 unstable angina,34 out-of-hospital resuscitation (with or without MI),1213 admissions for a variety of acute medical and surgical conditions,5556 and cardioversion from VT in the electrophysiology laboratory.62 Interestingly, epinephrine and norepinephrine rise postcardioversion from electrically induced VT, even of < 1 min in duration.63A canine model confirmed that LK post-resuscitation from VF was due to an acute serum potassium shift.64 Since LK is more common in patients with MI and a cardiac arrest than in those without arrest, the stress of the arrhythmic complication and/or the resuscitation appears to be a more powerful mediator of LK than MI, per se.1213

Direct documentation that LK in patients with MI or VF is due to a transient intracellular serum potassium shift has been provided in a few cases with serum potassium measurements shortly before and after the morbid event, the former obtained fortuitously.62,6566 Although some patients receive sodium bicarbonate during resuscitation, the resulting mild metabolic alkalosis does not affect serum potassium levels and is not the mechanism of LK.64,67

Indirect documentation for a serum potassium shift-dependent LK is suggested by the following: (1) the finding of LK after out-of-hospital resuscitation in up to one half of patients, all of whom were not taking diuretics1213; (2) the protective effect for LK development of noncardioselective andβ 2-blockade,3,36,51; (3) the nonemergence of LK with administration ofβ 1-adrenergic agonists,5152; (4) the rise of catecholamines early in MI50; (5) the occurrence of LK in patients seen early in the clinical course of MI, noted in our patients; (6) the rapid correction of LK with small amounts of serum potassium supplementation, suggestive of a reverse outward serum potassium cellular shift during the poststress period12; and (7) the spontaneous rise of low initial serum potassium levels in patients admitted to the hospital, irrespective of serum potassium replacement.56

Frequency of LK in MI

Varying reported rates of LK in MI may be due to differences in the threshold for diagnosis of LK (ranging from < 3.5 to < 4.0 mEq/L),11,16,23,35 the timing of serum potassium measurement (extended from emergency department admission to up to 3 days thereafter),23,35 or the contribution of prior diuretic therapy.8,14 Like others,8,1112,23 we did not find a relation between prior diuretic use and LK, although some have detected such an effect.10,25,35 A more severe“ stress-induced” LK in patients with MI who have been taking diuretics is suggested by the findings from normal volunteers who showed a synergistic LK effect from epinephrine infusions and prior use of diuretics58; the serum potassium drop in the latter was similar to the one noted in patients with MI.58 Over the last decade, the prevailing underutilization of diuretics in the management of hypertension could have contributed to the somewhat decreased rate of LK in our patients. Only 24.8% of our patients used diuretics prior to admission, a much lower rate than noted in older studies.7 Conversely, the recent resurgence in popularity of diuretics68could lead again to a rise in the incidence of LK in patients with MI. The stress-mediating mechanism of LK in patients with MI does not exclude the infrequent occurrence of LK due primarily to diuretic-induced total body serum potassium depletion.69

Only 7 of our 41 patients (17.0%) with LK had serum potassium < 3.0 mEq/L, and none had serum potassium < 2.5 mEq/L, a commonly employed threshold for diagnosis of severe LK.27,29 In contrast to other studies,70 and in agreement with some workers,19 we did not find a correlation between emergency department serum potassium and magnesium measurements.

Serial Serum Potassium Measurements

An association has been posed in the literature between LK, as defined by a single measurement of serum potassium carried out at a varying time point, and an untoward in-hospital outcome of patients with MI.19,35 Intuitively, it appears more fitting to link complications to a number of serum potassium values, leading to an“ operational” serum potassium, exerting its effect over the course of hospitalization. At the study design stage, this was the motivation for the use of CCU and mean serum potassium levels in our study (Table 1). However, since arrhythmias occurred early in admission (Table 3), and the CCU and mean serum potassium levels of the LK patients were lower than the corresponding values of the NK cohort, it was thought to be redundant to also analyze the data according to the CCU and mean serum potassium values. Thus, it was shown that emergency department serum potassium levels can be used after all as a variable for outcome correlations.

A rise of serum potassium values from the emergency department to the CCU, and to the mean value was noted in the LK cohort. We could not document a spontaneous rise of serum potassium after transfer to the CCU, as we did with magnesium,37 since serum potassium supplementation commenced immediately after detection of LK, often in the emergency department. However, severe LK in patients after out-of-hospital resuscitation returned to normal within 14 to 16 h with only modest serum potassium supplementation, suggesting an out-of-cell serum potassium shift during recovery.12 Also LK in patients with hemodynamically significant VT of <1 min in duration induced in the electrophysiology laboratory and with normal baseline serum potassium has resolved spontaneously within 3 h.62

Why did serum potassium levels not rise eventually in the LK cohort to the levels noted in the NK patients? Does a persistently low serum potassium (albeit normal) level identify patients at a high degree of stress throughout hospitalization, destined to manifest persistently low serum potassium values? This is not, however, supported by heart rates on admission (a rough index of adrenergic surge) that were not different in the two cohorts (Table 1). In contrast, an unexplained significant drop of serum potassium was noted between the emergency department and the CCU in the NK patients. Could this imply that the stress-induced intracellular serum potassium shift continues to operate in the hours after admission, and that the rise of serum potassium is initially influenced mainly by the serum potassium supplementation (43.1% for NK patients vs 90.2% for LK patients; Table 2)? This could be only discerned by frequent serum potassium levels and catecholamine measurements, carried out during the acute phase of MI.

Characterization of LK and NK Cohorts

Earlier emergency department admission of the LK patients could have been the reason for their lower emergency department serum potassium levels (high catecholamine levels are found early in the MI course),48,50 or could have been merely due to their emergent transport following VF, which occurred more frequently in this group. Moreover, VA is more frequent in the early course of MI, with a precipitous subsequent drop in its rate.70 Thus, VA, the early phase of MI, and LK are probably interrelated.

In agreement with others,35 we found larger MIs in the LK patients. However, this cohort did not have lower ejection fraction, higher admission or peak Killip class, or more frequent anterior MI, pulmonary congestion, or cardiomegaly, attributes expected to be encountered in patients with larger MIs. In both experimental work and clinical studies, a good correlation has been shown between MI size and malignant VA44,7172; also, these two variables have been linked to high catecholamine levels in patients with MI.4850,73 Epinephrine, at levels commensurate with the ones noted in patients with MI, infused in normal subjects led to LK,51,53,74 and pretreatment with blockers prevented its emergence.25,5152,75 Finally, it is also possible that catecholamine release may be the cause of VA in patients with MI, with LK being a mere index of such metabolic surge.11 Thus, MI size, rate of VA, and degree of adrenergic arousal (with its resultant LK) appear to be interrelated.

Complications Attributed to LK in MI

Patients with MI and LK are not at risk for increased mortality,10,19 although there is one study23 reporting such a relation. Also the mortality issue for merely hypertensive patients receiving diuretics (with suspected or proven LK) led to controversy in the past,2728 although more recent data do not point to an increased risk from the mild or moderate LK engendered by diuretics.3,27,30,32,68

Only VF occurred more frequently in LK patients than in NK patients in our study; its rate was higher than noted previously,10,34 probably because we included incidences happening even prior to admission.36 Also, we have shown that only pre-emergency department VF occurred with greater frequency in the LK cohort, while VF in all other settings occurred at the same rate in both subgroups. VT and VF have been previously attributed to LK in patients with MI, particularly early in the clinical course,7,10,14,19,23 although not invariably.11,16

Experimental work suggests that chronic LK, per se, is arrhythmogenic, more so in conjunction with MI. Both the threshold for spontaneous or electrically induced VT and VF have decreased proportionally to the attained LK.72,7677 However, such studies may not be relevant to our investigation, since the effected LK was associated with a total body serum potassium deficit (induced by diet and diuretics, or hemodialysis)72,77; also, these animal models employed severe LK, as compared to the serum potassium values detected in clinical settings.72,77 None of our patients had serum potassium levels < 2.5 mEq/L, a value considered to be the threshold for diagnosis of severe LK.27,29,68 We did not find a different rate of VT in the two cohorts, but others have reported VT to be inversely related to serum potassium.19

Large MIs and the early phase of the illness, both known to be independently linked to high rate of VA,7,19,23,44,72 inevitably have confounded attempts to associate causal VA with LK. Thus, VF was found more frequently in association with LK, and it occurred early in the clinical course.10 Was VT due to LK present on admission? Was VT related to the catecholamine levels at the time? Was it due to the interplay of these, or more factors? These issues have not been unraveled yet.

Correlations of PVCs with serum potassium values in the literature have been controversial, with some studies reporting an inverse relation,17 while others finding none.7 Some have detected a relationship between the initial serum potassium and PVCs in Holter recordings from the first 12 h after admission.19 In our study, the rate of PVCs was the same in the two subgroups. The work on LK in patients taking diuretics, in whom the rate of PVCs correlated with serum potassium levels,2 does not apply to the current practice of using low doses of diuretics,30 or to patients with MI, with transient LK, unassociated with total body serum potassium loss.

We documented similar rates of atrial fibrillation in our two cohorts, while LK has been previously implicated as a predisposing factor for such an arrhythmia.78

Matters Pertaining to Therapy

The increased rate of VA found in our patients with LK early on presentation could have led to an excessive use of antiarrhythmics, thus inevitably confounding the study of the independent arrhythmogenic role of LK. Thus, the notion is inescapable that had our patients with LK not received serum potassium and magnesium supplements and antiarrhythmics at a higher rate than the NK patients, they would have experienced even more frequent VA than was detected. Serum potassium and magnesium supplementation and antiarrhythmics were withheld for 12 h in a previous study, but this occurred in an environment where the patients had continuous ECG recordings for only 12 h; this study detected an inverse relationship of LK with self-limiting VT and PVCs.19 VA is seen early in the MI clinical course6,10,19; therefore, serum potassium maintenance at normal values through supplementation may be particularly important during this time frame. For the same reason, preadmission use of noncardioselective β-blockers is expected to provide amelioration or full protection from incipient LK.34 The relationship of LK and various types of VA can only be assessed by observing patients throughout hospitalization without correcting LK or resorting to the use of antiarrythmics, both unacceptable premises.

IV nitroglycerin was used less often in our patients with LK (Table 2), and this could be potentially traced to the lower systolic and diastolic BPs.

The lower rate of use of heparin in the LK cohort (Table 2) was probably due to a higher threshold for employing such therapy in patients who had undergone resuscitation more frequently (Table 3).

The effect of β-blockers in offsetting LK could not be studied in our patients, since only a small number (similar in the two subgroups) of patients were using such drugs prior to admission (Table 1).) Also, others have not found an effect of such therapy on serum potassium in patients with MI.11 Early presentation to emergency department and large MIs, which both promote high catecholamine levels, are expected to mitigate some of the protective effects for the development of LK, and it is probably the reason that β-blockers do not completely prevent LK, while they may still prevent VA.25

Study Limitations

The nonavailability of serum potassium values prior to the occurrence of VT and VF, and the confounding effect of serum potassium supplementation constitute the limitations of our work on strictly scientific grounds. However, there was no way to provide for premorbid serum potassium measures in a study of patients admitted with MI. Also, an established therapy such as serum potassium supplementation for patients with LK could not be withheld.11

The lower rate of thrombolysis in our patients (26.2%) than noted in a national registry (39%)79 might be thought to prevent the extrapolation of our findings to the current era; however, this therapy was applied equally to the LK and NK groups, and thus the objective of our study to evaluate the independent role of LK in MI was not compromised.

Clinical Implications

The association of LK with early presentation to the emergency department or with a large MI may alert the clinician about the acuteness and severity of the patient’s illness, since these two attributes (linked to high catecholamine values) constitute a substrate for emergence of complications.

LK is also found in patients presenting to the emergency department with a host of other potentially catastrophic illnesses,56 and thus it may be viewed as an index of the acuteness or stress imparted by many clinical conditions. In this respect, LK seems to be a consequence to a stereotyped physiologic response to stress. Since 7.9% of our patients had LK, while 25.9% had hypomagnesemia,37 and such a difference was also noted in previous studies,4,7,20,73,80 LK may represent a less sensitive indicator of stress than hypomagnesemia.37

Abbreviations: CCU = coronary care unit; CK = creatine kinase; CK-MB = myocardial isomer of CK; LK = hypokalemia; MI = myocardial infarction; NK = normokalemia; PVCs = premature ventricular contractions; VA = ventricular arrhythmia; VF = ventricular fibrillation; VT = ventricular tachycardia

Table Graphic Jump Location
Table 1. Demographic, Clinical, and Laboratory Characteristics of Patients With MI According to Stratification in LK and NK Groups Based on Emergency Department Serum Potassium Values*
* 

Data are presented as mean ± SD unless otherwise indicated. ED = emergency department; SBP = systolic BP; DBP = diastolic BP; CAD = coronary artery disease; F = female; M = male.

 

Forty LK patients and 472 NK patients had this measurement.

 

Seventeen LK patients and 138 NK patients had this measurement.

§ 

Thirteen LK patients and 124 NK patients had this measurement.

 

Ejection fraction was measured in 46.3% of patients with LK and 44.7% of patients with NK.

 

Angiography was done in 43.9% of patients with LK and 49.6% of patients with NK.

Table Graphic Jump Location
Table 2. In-Hospital Management of Patients With MI According to Stratification in LK and NK Groups Based on the Emergency Department Serum Potassium Values*
* 

Data are presented as %.

Table Graphic Jump Location
Table 3. Clinical Outcome of Patients With MI According to Stratification in LK and NK Groups Based on the Emergency Department Serum Potassium Values*
* 

Data are presented as % unless otherwise indicated. See Table 1 for abbreviation.

 

The difference in the rate of occurrence of VF in the pre-ED setting in the LK and NK cohorts was the main contributor to the statistically significant χ2 test result, as assessed by the observed and expected frequencies of VF in the two groups, and in the four different locations where it was detected.

Holland, OB, Nixon, JV, Kuhnert, L (1981) Diuretic-induced ventricular ectopic activity.Am J Med78,762-768
 
Hollifield, JW, Slaton, PE Thiazide diuretics, hypokalemia and cardiac arrhythmias.Acta Med Scand Suppl1981;647,67-73. [PubMed]
 
Papademetriou, V, Burris, JF, Notargiacomo, A, et al Thiazide therapy is not a cause of arrhythmia in patients with systemic hypertension.Arch Intern Med1988;148,1272-1276. [CrossRef] [PubMed]
 
Madias, JE, Madias, NE, Gavras, HP Nonarrhythmogenicity of diuretic-induced hypokalemia: its evidence in patients with uncomplicated hypertension.Arch Intern Med1984;144,2171-2176. [CrossRef] [PubMed]
 
Dyckner, T, Halmers, C, Lundman, T, et al Initial serum potassium level in relation to early complications and prognosis in patients with acute myocardial infarction.Acta Med Scand1975;197,207-210. [PubMed]
 
Bech, OA, Hochrein, H Initial serum potassium level in relation to cardiac arrhythmias in acute myocardial infarction.Z Kardiol1977;66,187-190. [PubMed]
 
Solomon, RJ, Cole, AG Importance of potassium in patients with acute myocardial infarction.Acta Med Scand Suppl1981;647,87-93. [PubMed]
 
Nodrehaug, JE Malignant arrhythmias in relation to serum potassium values in patients with an acute myocardial infarction.Acta Med Scand Suppl1981;647,101-107. [PubMed]
 
Brezins, M, Elyassov, S, Elimelech, I, et al Comparison of patients with acute myocardial infarction with and without ventricular fibrillation.Am J Cardiol1996;78,948-950. [CrossRef] [PubMed]
 
Nordrehaug, JE, Von der Lippe, G Hypokalemia and ventricular fibrillation in acute myocardial infarction.Br Heart J1983;50,525-529. [CrossRef] [PubMed]
 
Kafka, H, Langevin, L, Armstrong, PW Serum magnesium and potassium in acute myocardial infarction: influence on ventricular arrhythmias.Arch Intern Med1987;147,465-469. [CrossRef] [PubMed]
 
Salerno, DM, Asinger, RW, Elsperger, J, et al Frequency of hypokalemia after successfully resuscitated out-of-hospital acute myocardial infarction.Am J Cardiol1987;59,84-88. [CrossRef] [PubMed]
 
Thompson, G, Cobb, LA Hypokalemia after resuscitation from out-of-hospital ventricular fibrillation.JAMA1982;248,2860-2863. [CrossRef] [PubMed]
 
Duke, M Thiazide-induced hypokalemia: association with acute myocardial infarction and ventricular fibrillation.JAMA1978;239,43-45. [CrossRef] [PubMed]
 
Reuben, SR, Thomas, RD The relationship between serum potassium and cardiac arrhythmias following cardiac infarction in patients aged over 65 years.Curr Med Res Opin1982;7(suppl 1),79-82
 
Rao, SK The arrhythmic danger of hypokalemia.Curr Med Res Opin1982;7(suppl 1),83-86
 
Cooper, WD, Kwan, P, Reuben, SR, et al Cardiac arrhythmias following acute myocardial infarction: association with the serum potassium level and prior diuretic therapy.Eur Heart J1984;6,464-469
 
Nordrehaug, JE Malignant arrhythmia in relation to serum potassium in acute myocardial infarction.Am J Cardiol1985;56,20D-23D. [CrossRef] [PubMed]
 
Nordrehaug, JE, Johannessen, K-A, Von Der Lippe, G Serum potassium concentration as a risk factor of ventricular arrhythmias early in acute myocardial infarction.Circulation1985;71,645-649. [CrossRef] [PubMed]
 
Abraham, AS, Rosenman, D, Meshulam, Z, et al Serum, lymphocyte, and erythrocyte potassium, magnesium, and calcium concentrations and their relation to tachyarrythmias in patients with acute myocardial infarction.Am J Med1986;81,983-988. [CrossRef] [PubMed]
 
Clausen, TG, Brocks, K, Ibsen, H Hypokalemia and ventricular arrhythmias in acute myocardial infarction.Acta Med Scand1988;224,531-537. [PubMed]
 
Podger, JC, Simpson, E, Rolton, HA, et al The hypokalemia of acute myocardial infarction.Ann Clin Biochem1986;23,204-205. [PubMed]
 
Thomas, RD Ventricular fibrillation and initial plasma potassium in acute myocardial infarction.Postgrad Med J1983;59,354-356. [CrossRef] [PubMed]
 
Dyckner, T, Helmers, C, Wester, PO Initial serum potassium level, early arrhythmias and previous diuretic therapy in acute myocardial infarction.Int J Cardiol1982;2,146-148. [CrossRef] [PubMed]
 
Johansson, BW, Dziamski, R Malignant arrhythmias in acute myocardial infarction: relationship to serum potassium and effect of selective and non-selective β-blockade.Drugs1984;28(suppl I),77-85
 
Hulting, J In-hospital ventricular fibrillation and its relation to serum potassium.Acta Med Scand Suppl1981;647,109-116. [PubMed]
 
Harrington, JT, Isner, JM, Kassirer, JP Our national obsession with potassium.Am J Med1982;73,155-159
 
Kaplan, NM Our appropriate concern about hypokalemia.Am J Med1984;77,1-4
 
Kassirer, JP, Harrington, JT Fending off the potassium pushers [editorial].N Engl J Med1985;312,785-787. [CrossRef] [PubMed]
 
Moser, M Why are physicians not prescribing diuretics more frequently in the management of hypertension?JAMA1998;279,1813-1816. [CrossRef] [PubMed]
 
Freis, ED Critique of the clinical importance of diuretic-induced hypokalemia and elevated cholesterol level.Arch Intern Med1989;149,2640-2648. [CrossRef] [PubMed]
 
Multiple Risk Factor Intervention Trial Research Group.. Baseline rest electrocardiographic abnormalities, antihypertensive treatment, and mortality in the multiple risk factor intervention trial.Am J Cardiol1985;55,1-15. [CrossRef] [PubMed]
 
The Hypertension Detection and Follow-up Program Cooperative Research Group. The effect of antihypertensive drug treatment on mortality in the presence of resting electrocardiographic abnormalities at baseline: the HDFP experience.Circulation1984;70,996-1003. [CrossRef] [PubMed]
 
Higham, PD, Adams, PC, Murray, A, et al Plasma potassium, serum magnesium and ventricular fibrillation: a prospective study.QJM1993;86,609-617. [PubMed]
 
Herlitz, J, Hjalmarson, A, Bengtson, A Occurrence of hypokalemia in suspected acute myocardial infarction and its relation to clinical history and clinical course.Clin Cardiol1988;11,678-682. [CrossRef] [PubMed]
 
Nordrehaug, JE, Johannessen, KA, von der Lippe, G, et al Effect of timolol on changes in serum potassium concentration during acute myocardial infarction.Br Heart J1985;53,388-393. [CrossRef] [PubMed]
 
Madias, JE, Sheth, K, Choudry, MA, et al Admission serum magnesium level does not predict the hospital outcome of patients with acute myocardial infarction.Arch Intern Med1996;156,1701-1708. [CrossRef] [PubMed]
 
Madias, JE, Chahine, RA, Gorlin, R, et al A comparison of transmural and nontransmural acute myocardial infarction.Circulation1974;49,498-507. [CrossRef] [PubMed]
 
Cooperating. investigators from the MILIS study group: electrocardiographic, enzymatic and scintigraphic criteria of acute myocardial infarction as determined from study of 726 patients (MILIS study). Am J Cardiol l985; 55:1463–1468.
 
Killip, T, III, Kimball, JT Treatment of myocardial infarction in a coronary care unit: a two year experience with 250 patients.Am J Cardiol1967;20,457-464. [CrossRef] [PubMed]
 
Tietz, NW Fundamentals of clinical chemistry.1983,963 W.B. Saunders. Philadelphia, PA:
 
Tietz, NW Textbook of clinical chemistry.1986,974 W.B. Saunders. Philadelphia, PA:
 
Rentrop, KP, Feit, F, Sherman, W, et al Late thrombolytic therapy preserves left ventricular function in patients with collateralized total coronary occlusion: primary end point findings of the second Mount Sinai-New York University reperfusion trial.J Am Coll Cardiol1989;14,58-64. [CrossRef] [PubMed]
 
Cox, JR, Jr, Roberts, R, Ambos, HD, et al Relations between enzymatically estimated myocardial infarct size and early ventricular dysrhythmia.Circulation1976;53(Suppl I),150-155
 
Colton T. Statistics in medicine. 1st ed. Boston, MA: Little, Brown and Company, 1974;179, 210.
 
Norusis M. The SPSS guide to data analysis for SPSS/SPC+. 2nd ed. Chicago, IL: SPSS, 1991; 88, 218, 231, 244, 268.
 
Norusis M. SPSS/PC+4.0 statistics. Chicago, IL: SPSS, 1991; B-15.
 
Valori, C, Thomas, M, Shillingford, J Free noradrenaline and adrenaline excretion in relation to clinical syndromes following myocardial infarction.Am J Cardiol1967;20,605-617. [CrossRef] [PubMed]
 
Jewitt, DE, Mercer, CJ, Reid, D, et al Free noradrenaline and adrenaline excretion in relation to the development of cardiac arrhythmias and heart failure in patients with acute myocardial infarction.Lancet1969;1,635-641. [PubMed]
 
Karlsberg, RP, Cryer, PE, Roberts, R Serial plasma catecholamine response early in the course of clinical acute myocardial infarction: relationship to infarct extent and mortality.Am Heart J1981;102,24-29. [CrossRef] [PubMed]
 
Brown, MJ, Brown, DC, Murphy, MB Hypokalemia from beta2-receptor stimulation by circulating epinephrine.N Engl J Med1983;309,1414-1419. [CrossRef] [PubMed]
 
Vincent HH, Boomsma F, Man in’t Veld AJ, et al. Effects of selective and nonselective b-agonists on plasma potassium and norepinephrine J Cardiovasc Pharmacol 1984; 6:107–114.
 
Struthers, AD, Reid, JL, Whitesmith, R, et al Effects of intravenous adrenaline on electrocardiogram, blood pressure, and serum potassium.Br Heart J1983;49,90-93. [CrossRef] [PubMed]
 
Clausen, T, Flatman, JA β2-adrenoceptors mediate the stimulating effect of adrenaline on active electrogenic Na- K transport in rat soleus muscle.Br J Pharmacol1980;68,749-755. [CrossRef] [PubMed]
 
Smith, JS, Jr Hypokalemia in resuscitation from multiple trauma.Surg Gynecol Obstet1978;147,18-20. [PubMed]
 
Morgan, DB, Young, RM Acute transient hypokalemia: new interpretation of a common event.Lancet1982;2,751-752. [PubMed]
 
Todd, EP, Vick, RL Kalemotropic effect of epinephrine: analysis with adrenergic agonists and antagonists.Am J Physiol1971;220,1964-1969. [PubMed]
 
Struthers, AD, Whitesmith, R, Reid, JL Prior thiazide diuretic treatment increases adrenaline-induced hypokalemia.Lancet1983;1,1358-1361. [PubMed]
 
Gelmont, DM, Balmes, JR, Yee, A Hypokalemia induced by inhaled bronchodilators.Chest1988;94,763-766. [CrossRef] [PubMed]
 
Liou, H-H, Chiang, S-S, Wu, S-C, et al Hypokalemic effects of intravenous infusion or nebulization of salbutamol in patients with chronic renal failure: comparative study.Am J Kidney Dis1994;23,266-271. [PubMed]
 
Allon, M, Copkney, C Albuterol and insulin for treatment of hyperkalemia in hemodialysis patients.Kidney Int1990;38,869-872. [CrossRef] [PubMed]
 
Salerno, DM, Dunbar, D, Sharkey, P Hypokalemia after cardioversion from ventricular tachycardia induced in the electrophysiology laboratory.Am Heart J1987;114,1389-1395. [CrossRef] [PubMed]
 
Morady, F, DiCarlo, LA, Halter, JB, et al The plasma catecholamine response to ventricular tachycardia induction and external countershock during electrophysiologic testing.J Am Coll Cardiol1986;8,584-591. [CrossRef] [PubMed]
 
Salerno, DM, Elsperger, J, Helseth, P, et al Serum potassium, calcium, and magnesium after resuscitation from ventricular fibrillation: a canine study.J Am Coll Cardiol1987;10,178-185. [CrossRef] [PubMed]
 
Salerno, DM Postrescuscitation hypokalemia in a patient with a normal prearrest serum potassium level.Ann Intern Med1988;108,836-837. [PubMed]
 
Isner, JM, Harten, JT Factitious lowering of the serum potassium level after cardiopulmonary resuscitation. Implications for evaluating the arrhythmogenicity of hypokalemia in acute myocardial infarction.Arch Intern Med1985;145,161-162. [CrossRef] [PubMed]
 
Adrogue, HJ, Madias, NE Changes in plasma potassium concentration during acute acid-base disturbances.Am J Med1981;71,456-467. [CrossRef] [PubMed]
 
The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Arch Intern Med 1997; 157:2413–2446.
 
Ruder, MA, Flaker, GC, Alpert, MA, et al Hypokalemia as a cause of cardiac arrest: results of electrophysiologic testing and long-term follow-up.Am Heart J1985;110,490-491. [CrossRef] [PubMed]
 
Dyckner, T Serum magnesium in acute myocardial infarction: relation to arrhythmias.Acta Med Scand1980;207,59-66. [PubMed]
 
Meltzer, LE, Cohen, HE The incidence of arrhythmias associated with acute myocardial infarction. Meltzer, LE Dunning, AJ eds.Textbook of coronary care1972,107-133 Charles Press. Philadelphia, PA:
 
Garan, H, McGovern, BA, Canzanello, VJ, et al The effect of potassium ion depletion on postinfarction canine cardiac arrhythmias.Circulation1988;77,696-704. [CrossRef] [PubMed]
 
Vetter, NJ, Adams, W, Strange, RC, et al Initial metabolic and hormonal response to acute myocardial infarction.Lancet1974;1,284-289. [PubMed]
 
Brown, MJ Hypokalemia from β2-receptor stimulation by circulating epinephrine.Am J Cardiol1985;56,3D-9D. [CrossRef] [PubMed]
 
Simpson, E, Rodger, JC, Raj, SM, et al Pre-treatment with β-blockers and the frequency of hypokalemia in patients with acute chest pain.Br Heart J1987;58,499-504. [CrossRef] [PubMed]
 
Yano, K, Hirata, M, Matsumoto, Y, et al Effects of chronic hypokalemia on ventricular vulnerability during acute myocardial ischemia in the dog.Jpn Heart J1989;30,205-217. [CrossRef] [PubMed]
 
Hohnloser, SH, Verrier, RL, Lown, B, et al Effects of hypokalemia on susceptibility to ventricular fibrillation in the normal and ischemic canine heart.Am Heart J1986;112,32-35. [CrossRef] [PubMed]
 
Olson, SB Nature of cardiac arrhythmias and electrolyte disturbances: role of potassium in atrial fibrillation.Acta Med Scand Suppl1981;647,33-37. [PubMed]
 
National Heart Attack Alert Program Coordinating Committee, 60 Minutes to Treatment Working Group. Emergency department: rapid identification and treatment of patients with acute myocardial infarction.Ann Emerg Med1994;23,311-329. [CrossRef] [PubMed]
 
Haffner, CA, Kendall, MJ Metabolic effects of β2-agonists.J Clin Pharm Ther1992;17,155-164. [CrossRef] [PubMed]
 

Figures

Tables

Table Graphic Jump Location
Table 1. Demographic, Clinical, and Laboratory Characteristics of Patients With MI According to Stratification in LK and NK Groups Based on Emergency Department Serum Potassium Values*
* 

Data are presented as mean ± SD unless otherwise indicated. ED = emergency department; SBP = systolic BP; DBP = diastolic BP; CAD = coronary artery disease; F = female; M = male.

 

Forty LK patients and 472 NK patients had this measurement.

 

Seventeen LK patients and 138 NK patients had this measurement.

§ 

Thirteen LK patients and 124 NK patients had this measurement.

 

Ejection fraction was measured in 46.3% of patients with LK and 44.7% of patients with NK.

 

Angiography was done in 43.9% of patients with LK and 49.6% of patients with NK.

Table Graphic Jump Location
Table 2. In-Hospital Management of Patients With MI According to Stratification in LK and NK Groups Based on the Emergency Department Serum Potassium Values*
* 

Data are presented as %.

Table Graphic Jump Location
Table 3. Clinical Outcome of Patients With MI According to Stratification in LK and NK Groups Based on the Emergency Department Serum Potassium Values*
* 

Data are presented as % unless otherwise indicated. See Table 1 for abbreviation.

 

The difference in the rate of occurrence of VF in the pre-ED setting in the LK and NK cohorts was the main contributor to the statistically significant χ2 test result, as assessed by the observed and expected frequencies of VF in the two groups, and in the four different locations where it was detected.

References

Holland, OB, Nixon, JV, Kuhnert, L (1981) Diuretic-induced ventricular ectopic activity.Am J Med78,762-768
 
Hollifield, JW, Slaton, PE Thiazide diuretics, hypokalemia and cardiac arrhythmias.Acta Med Scand Suppl1981;647,67-73. [PubMed]
 
Papademetriou, V, Burris, JF, Notargiacomo, A, et al Thiazide therapy is not a cause of arrhythmia in patients with systemic hypertension.Arch Intern Med1988;148,1272-1276. [CrossRef] [PubMed]
 
Madias, JE, Madias, NE, Gavras, HP Nonarrhythmogenicity of diuretic-induced hypokalemia: its evidence in patients with uncomplicated hypertension.Arch Intern Med1984;144,2171-2176. [CrossRef] [PubMed]
 
Dyckner, T, Halmers, C, Lundman, T, et al Initial serum potassium level in relation to early complications and prognosis in patients with acute myocardial infarction.Acta Med Scand1975;197,207-210. [PubMed]
 
Bech, OA, Hochrein, H Initial serum potassium level in relation to cardiac arrhythmias in acute myocardial infarction.Z Kardiol1977;66,187-190. [PubMed]
 
Solomon, RJ, Cole, AG Importance of potassium in patients with acute myocardial infarction.Acta Med Scand Suppl1981;647,87-93. [PubMed]
 
Nodrehaug, JE Malignant arrhythmias in relation to serum potassium values in patients with an acute myocardial infarction.Acta Med Scand Suppl1981;647,101-107. [PubMed]
 
Brezins, M, Elyassov, S, Elimelech, I, et al Comparison of patients with acute myocardial infarction with and without ventricular fibrillation.Am J Cardiol1996;78,948-950. [CrossRef] [PubMed]
 
Nordrehaug, JE, Von der Lippe, G Hypokalemia and ventricular fibrillation in acute myocardial infarction.Br Heart J1983;50,525-529. [CrossRef] [PubMed]
 
Kafka, H, Langevin, L, Armstrong, PW Serum magnesium and potassium in acute myocardial infarction: influence on ventricular arrhythmias.Arch Intern Med1987;147,465-469. [CrossRef] [PubMed]
 
Salerno, DM, Asinger, RW, Elsperger, J, et al Frequency of hypokalemia after successfully resuscitated out-of-hospital acute myocardial infarction.Am J Cardiol1987;59,84-88. [CrossRef] [PubMed]
 
Thompson, G, Cobb, LA Hypokalemia after resuscitation from out-of-hospital ventricular fibrillation.JAMA1982;248,2860-2863. [CrossRef] [PubMed]
 
Duke, M Thiazide-induced hypokalemia: association with acute myocardial infarction and ventricular fibrillation.JAMA1978;239,43-45. [CrossRef] [PubMed]
 
Reuben, SR, Thomas, RD The relationship between serum potassium and cardiac arrhythmias following cardiac infarction in patients aged over 65 years.Curr Med Res Opin1982;7(suppl 1),79-82
 
Rao, SK The arrhythmic danger of hypokalemia.Curr Med Res Opin1982;7(suppl 1),83-86
 
Cooper, WD, Kwan, P, Reuben, SR, et al Cardiac arrhythmias following acute myocardial infarction: association with the serum potassium level and prior diuretic therapy.Eur Heart J1984;6,464-469
 
Nordrehaug, JE Malignant arrhythmia in relation to serum potassium in acute myocardial infarction.Am J Cardiol1985;56,20D-23D. [CrossRef] [PubMed]
 
Nordrehaug, JE, Johannessen, K-A, Von Der Lippe, G Serum potassium concentration as a risk factor of ventricular arrhythmias early in acute myocardial infarction.Circulation1985;71,645-649. [CrossRef] [PubMed]
 
Abraham, AS, Rosenman, D, Meshulam, Z, et al Serum, lymphocyte, and erythrocyte potassium, magnesium, and calcium concentrations and their relation to tachyarrythmias in patients with acute myocardial infarction.Am J Med1986;81,983-988. [CrossRef] [PubMed]
 
Clausen, TG, Brocks, K, Ibsen, H Hypokalemia and ventricular arrhythmias in acute myocardial infarction.Acta Med Scand1988;224,531-537. [PubMed]
 
Podger, JC, Simpson, E, Rolton, HA, et al The hypokalemia of acute myocardial infarction.Ann Clin Biochem1986;23,204-205. [PubMed]
 
Thomas, RD Ventricular fibrillation and initial plasma potassium in acute myocardial infarction.Postgrad Med J1983;59,354-356. [CrossRef] [PubMed]
 
Dyckner, T, Helmers, C, Wester, PO Initial serum potassium level, early arrhythmias and previous diuretic therapy in acute myocardial infarction.Int J Cardiol1982;2,146-148. [CrossRef] [PubMed]
 
Johansson, BW, Dziamski, R Malignant arrhythmias in acute myocardial infarction: relationship to serum potassium and effect of selective and non-selective β-blockade.Drugs1984;28(suppl I),77-85
 
Hulting, J In-hospital ventricular fibrillation and its relation to serum potassium.Acta Med Scand Suppl1981;647,109-116. [PubMed]
 
Harrington, JT, Isner, JM, Kassirer, JP Our national obsession with potassium.Am J Med1982;73,155-159
 
Kaplan, NM Our appropriate concern about hypokalemia.Am J Med1984;77,1-4
 
Kassirer, JP, Harrington, JT Fending off the potassium pushers [editorial].N Engl J Med1985;312,785-787. [CrossRef] [PubMed]
 
Moser, M Why are physicians not prescribing diuretics more frequently in the management of hypertension?JAMA1998;279,1813-1816. [CrossRef] [PubMed]
 
Freis, ED Critique of the clinical importance of diuretic-induced hypokalemia and elevated cholesterol level.Arch Intern Med1989;149,2640-2648. [CrossRef] [PubMed]
 
Multiple Risk Factor Intervention Trial Research Group.. Baseline rest electrocardiographic abnormalities, antihypertensive treatment, and mortality in the multiple risk factor intervention trial.Am J Cardiol1985;55,1-15. [CrossRef] [PubMed]
 
The Hypertension Detection and Follow-up Program Cooperative Research Group. The effect of antihypertensive drug treatment on mortality in the presence of resting electrocardiographic abnormalities at baseline: the HDFP experience.Circulation1984;70,996-1003. [CrossRef] [PubMed]
 
Higham, PD, Adams, PC, Murray, A, et al Plasma potassium, serum magnesium and ventricular fibrillation: a prospective study.QJM1993;86,609-617. [PubMed]
 
Herlitz, J, Hjalmarson, A, Bengtson, A Occurrence of hypokalemia in suspected acute myocardial infarction and its relation to clinical history and clinical course.Clin Cardiol1988;11,678-682. [CrossRef] [PubMed]
 
Nordrehaug, JE, Johannessen, KA, von der Lippe, G, et al Effect of timolol on changes in serum potassium concentration during acute myocardial infarction.Br Heart J1985;53,388-393. [CrossRef] [PubMed]
 
Madias, JE, Sheth, K, Choudry, MA, et al Admission serum magnesium level does not predict the hospital outcome of patients with acute myocardial infarction.Arch Intern Med1996;156,1701-1708. [CrossRef] [PubMed]
 
Madias, JE, Chahine, RA, Gorlin, R, et al A comparison of transmural and nontransmural acute myocardial infarction.Circulation1974;49,498-507. [CrossRef] [PubMed]
 
Cooperating. investigators from the MILIS study group: electrocardiographic, enzymatic and scintigraphic criteria of acute myocardial infarction as determined from study of 726 patients (MILIS study). Am J Cardiol l985; 55:1463–1468.
 
Killip, T, III, Kimball, JT Treatment of myocardial infarction in a coronary care unit: a two year experience with 250 patients.Am J Cardiol1967;20,457-464. [CrossRef] [PubMed]
 
Tietz, NW Fundamentals of clinical chemistry.1983,963 W.B. Saunders. Philadelphia, PA:
 
Tietz, NW Textbook of clinical chemistry.1986,974 W.B. Saunders. Philadelphia, PA:
 
Rentrop, KP, Feit, F, Sherman, W, et al Late thrombolytic therapy preserves left ventricular function in patients with collateralized total coronary occlusion: primary end point findings of the second Mount Sinai-New York University reperfusion trial.J Am Coll Cardiol1989;14,58-64. [CrossRef] [PubMed]
 
Cox, JR, Jr, Roberts, R, Ambos, HD, et al Relations between enzymatically estimated myocardial infarct size and early ventricular dysrhythmia.Circulation1976;53(Suppl I),150-155
 
Colton T. Statistics in medicine. 1st ed. Boston, MA: Little, Brown and Company, 1974;179, 210.
 
Norusis M. The SPSS guide to data analysis for SPSS/SPC+. 2nd ed. Chicago, IL: SPSS, 1991; 88, 218, 231, 244, 268.
 
Norusis M. SPSS/PC+4.0 statistics. Chicago, IL: SPSS, 1991; B-15.
 
Valori, C, Thomas, M, Shillingford, J Free noradrenaline and adrenaline excretion in relation to clinical syndromes following myocardial infarction.Am J Cardiol1967;20,605-617. [CrossRef] [PubMed]
 
Jewitt, DE, Mercer, CJ, Reid, D, et al Free noradrenaline and adrenaline excretion in relation to the development of cardiac arrhythmias and heart failure in patients with acute myocardial infarction.Lancet1969;1,635-641. [PubMed]
 
Karlsberg, RP, Cryer, PE, Roberts, R Serial plasma catecholamine response early in the course of clinical acute myocardial infarction: relationship to infarct extent and mortality.Am Heart J1981;102,24-29. [CrossRef] [PubMed]
 
Brown, MJ, Brown, DC, Murphy, MB Hypokalemia from beta2-receptor stimulation by circulating epinephrine.N Engl J Med1983;309,1414-1419. [CrossRef] [PubMed]
 
Vincent HH, Boomsma F, Man in’t Veld AJ, et al. Effects of selective and nonselective b-agonists on plasma potassium and norepinephrine J Cardiovasc Pharmacol 1984; 6:107–114.
 
Struthers, AD, Reid, JL, Whitesmith, R, et al Effects of intravenous adrenaline on electrocardiogram, blood pressure, and serum potassium.Br Heart J1983;49,90-93. [CrossRef] [PubMed]
 
Clausen, T, Flatman, JA β2-adrenoceptors mediate the stimulating effect of adrenaline on active electrogenic Na- K transport in rat soleus muscle.Br J Pharmacol1980;68,749-755. [CrossRef] [PubMed]
 
Smith, JS, Jr Hypokalemia in resuscitation from multiple trauma.Surg Gynecol Obstet1978;147,18-20. [PubMed]
 
Morgan, DB, Young, RM Acute transient hypokalemia: new interpretation of a common event.Lancet1982;2,751-752. [PubMed]
 
Todd, EP, Vick, RL Kalemotropic effect of epinephrine: analysis with adrenergic agonists and antagonists.Am J Physiol1971;220,1964-1969. [PubMed]
 
Struthers, AD, Whitesmith, R, Reid, JL Prior thiazide diuretic treatment increases adrenaline-induced hypokalemia.Lancet1983;1,1358-1361. [PubMed]
 
Gelmont, DM, Balmes, JR, Yee, A Hypokalemia induced by inhaled bronchodilators.Chest1988;94,763-766. [CrossRef] [PubMed]
 
Liou, H-H, Chiang, S-S, Wu, S-C, et al Hypokalemic effects of intravenous infusion or nebulization of salbutamol in patients with chronic renal failure: comparative study.Am J Kidney Dis1994;23,266-271. [PubMed]
 
Allon, M, Copkney, C Albuterol and insulin for treatment of hyperkalemia in hemodialysis patients.Kidney Int1990;38,869-872. [CrossRef] [PubMed]
 
Salerno, DM, Dunbar, D, Sharkey, P Hypokalemia after cardioversion from ventricular tachycardia induced in the electrophysiology laboratory.Am Heart J1987;114,1389-1395. [CrossRef] [PubMed]
 
Morady, F, DiCarlo, LA, Halter, JB, et al The plasma catecholamine response to ventricular tachycardia induction and external countershock during electrophysiologic testing.J Am Coll Cardiol1986;8,584-591. [CrossRef] [PubMed]
 
Salerno, DM, Elsperger, J, Helseth, P, et al Serum potassium, calcium, and magnesium after resuscitation from ventricular fibrillation: a canine study.J Am Coll Cardiol1987;10,178-185. [CrossRef] [PubMed]
 
Salerno, DM Postrescuscitation hypokalemia in a patient with a normal prearrest serum potassium level.Ann Intern Med1988;108,836-837. [PubMed]
 
Isner, JM, Harten, JT Factitious lowering of the serum potassium level after cardiopulmonary resuscitation. Implications for evaluating the arrhythmogenicity of hypokalemia in acute myocardial infarction.Arch Intern Med1985;145,161-162. [CrossRef] [PubMed]
 
Adrogue, HJ, Madias, NE Changes in plasma potassium concentration during acute acid-base disturbances.Am J Med1981;71,456-467. [CrossRef] [PubMed]
 
The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Arch Intern Med 1997; 157:2413–2446.
 
Ruder, MA, Flaker, GC, Alpert, MA, et al Hypokalemia as a cause of cardiac arrest: results of electrophysiologic testing and long-term follow-up.Am Heart J1985;110,490-491. [CrossRef] [PubMed]
 
Dyckner, T Serum magnesium in acute myocardial infarction: relation to arrhythmias.Acta Med Scand1980;207,59-66. [PubMed]
 
Meltzer, LE, Cohen, HE The incidence of arrhythmias associated with acute myocardial infarction. Meltzer, LE Dunning, AJ eds.Textbook of coronary care1972,107-133 Charles Press. Philadelphia, PA:
 
Garan, H, McGovern, BA, Canzanello, VJ, et al The effect of potassium ion depletion on postinfarction canine cardiac arrhythmias.Circulation1988;77,696-704. [CrossRef] [PubMed]
 
Vetter, NJ, Adams, W, Strange, RC, et al Initial metabolic and hormonal response to acute myocardial infarction.Lancet1974;1,284-289. [PubMed]
 
Brown, MJ Hypokalemia from β2-receptor stimulation by circulating epinephrine.Am J Cardiol1985;56,3D-9D. [CrossRef] [PubMed]
 
Simpson, E, Rodger, JC, Raj, SM, et al Pre-treatment with β-blockers and the frequency of hypokalemia in patients with acute chest pain.Br Heart J1987;58,499-504. [CrossRef] [PubMed]
 
Yano, K, Hirata, M, Matsumoto, Y, et al Effects of chronic hypokalemia on ventricular vulnerability during acute myocardial ischemia in the dog.Jpn Heart J1989;30,205-217. [CrossRef] [PubMed]
 
Hohnloser, SH, Verrier, RL, Lown, B, et al Effects of hypokalemia on susceptibility to ventricular fibrillation in the normal and ischemic canine heart.Am Heart J1986;112,32-35. [CrossRef] [PubMed]
 
Olson, SB Nature of cardiac arrhythmias and electrolyte disturbances: role of potassium in atrial fibrillation.Acta Med Scand Suppl1981;647,33-37. [PubMed]
 
National Heart Attack Alert Program Coordinating Committee, 60 Minutes to Treatment Working Group. Emergency department: rapid identification and treatment of patients with acute myocardial infarction.Ann Emerg Med1994;23,311-329. [CrossRef] [PubMed]
 
Haffner, CA, Kendall, MJ Metabolic effects of β2-agonists.J Clin Pharm Ther1992;17,155-164. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543