Author + information
- Received December 16, 1997
- Revision received March 2, 1998
- Accepted March 16, 1998
- Published online July 1, 1998.
- Mina K Chung, MD, FACCa,* (, )
- Robert A Schweikert, MDa,
- Bruce L Wilkoff, MD, FACCa,
- Mark J Niebauer, MDa,
- Sergio L Pinski, MDa,
- Richard G Trohman, MD, FACCa,
- Gregory A Kidwell, MD, FACCa,
- Fredrick J Jaeger, DO, FACCa,
- Victor A Morant, MD, FACCa,
- Dave P Miller, MS∗ and
- Patrick J Tchou, MD, FACCa
- ↵*Address for correspondence: Dr. Mina K. Chung, Department of Cardiology, Section of Electrophysiology and Pacing, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Desk F-15, Cleveland, Ohio 44195
Objectives. We sought to determine the yield of in-hospital monitoring for detection of significant arrhythmia complications in patients starting sotalol therapy for atrial arrhythmias and to identify factors that might predict safe outpatient initiation.
Background. The need for hospital admission during initiation of antiarrhythmic therapy has been questioned, particularly for sotalol, with which proarrhythmia may be dose related.
Methods. The records of 120 patients admitted to the hospital for initiation of sotalol therapy were retrospectively reviewed to determine the incidence of significant arrhythmia complications, defined as new or increased ventricular arrhythmias, significant bradycardia or excessive corrected QT (QTc) interval prolongation.
Results. Twenty-five patients (20.8%) experienced 35 complications, triggering therapy changes during the hospital period in 21 (17.5%). New or increased ventricular arrhythmias developed in 7 patients (5.8%) (torsade de pointes in 2), significant bradycardia in 20 (16.7%) (rate <40 beats/min in 13, pause >3.0 s in 4, third-degree atrioventricular block in 1, permanent pacemaker implantation in 3) and excessively prolonged QTc intervals in 8 (6.7%) (dosage reduced or discontinued in 6). Time to the earliest detection of complications was 2.1 ± 2.5 (mean ± SD) days after initiation of sotalol, with 22 of 25 patients meeting criteria for complications within 3 days of monitoring. Baseline electrocardiographic intervals or absence of heart disease failed to distinguish a low risk group. Multivariate analysis identified absence of a pacemaker as the only significant predictor of arrhythmia complications (p = 0.022).
Conclusions. Because clinically significant complications can be detected with in-hospital monitoring in one of five patients starting sotalol therapy, hospital admission is warranted for initiation of sotalol. Patients without pacemakers are at higher risk for these complications.
Hospital admission for monitoring during the initiation of primary antiarrhythmic drugs for atrial fibrillation remains controversial, particularly with sotalol. The risk of proarrhythmia from class I drugs has been reported to range from 1% to 8% (1–4). For sotalol, a class III potassium channel blocking agent that causes QT interval prolongation, this risk is 2% to 8% (5–8). In contrast to class IA antiarrhythmic agents, such as quinidine, with which proarrhythmia may be idiosyncratic (2,3,9–11), the proarrhythmia associated with sotalol may be more dose dependent (12,13), with risk potentially related to and detectable by monitoring of QT interval prolongation. As attention to health care cost containment and minimization of hospital lengths of stay has increased, the need for in-hospital initiation of sotalol has been questioned. However, sotalol does have the potential to cause significant life-threatening proarrhythmia, including significant bradycardia, which can enhance the risk of torsade de pointes, particularly in view of the reverse use dependence properties of sotalol (14,15).
In the current study we sought to determine 1) the yield of in-hospital monitoring for the detection of significant arrhythmias that could result in the need for intervention or dosage modification during the initiation of sotalol; and 2) whether baseline clinical, electrocardiographic (ECG) or echocardiographic characteristics could be used to predict which patients would be at particularly high risk for significant arrhythmia complications that would justify hospital admission and in which patients sotalol therapy might be safely initiated on an outpatient basis.
The hospital records of patients treated with sotalol at The Cleveland Clinic Foundation during the period from April 1993 through December 1993 were retrospectively reviewed. From pharmacy records, oral sotalol was dispensed to 174 patients during the study period. During the time period of the study, it was the policy of the cardiac electrophysiology section to initiate sotalol therapy in all patients in the hospital during telemetry with continuous recording. Record review identified 120 patients who received the drug for the first time for treatment of atrial arrhythmias during the index hospital period. These patients represented the focus of the present study. Patients were excluded from the study if sotalol was initiated primarily for the treatment of ventricular arrhythmias (n = 38), or if they had previously received sotalol (n = 13). Chart review revealed three patients who were not actually receiving sotalol. Baseline demographic, clinical, ECG and echocardiographic variables were collected. The incidence of significant arrhythmia complications and subsequent changes in therapies were determined from review of hospital records, including vital signs, telemetry or 24-h Holter monitoring and all ECGs. Telephone or chart follow-up, or both, was performed in all 104 patients discharged with sotalol (mean follow-up 7.8 months, range 4 days to 16.9 months).
Atrial arrhythmias were defined as chronicif they were persistent and present for ≥2 months before admission. Arrhythmias were classified as paroxysmalif they were intermittent or, if present on admission, had been sustained for <2 months before the index hospital admission. ECG intervals were determined using computer-generated measurements (Marquette Electronics, Inc.) and reviewed for accuracy by the investigators. The corrected QT interval (QTc) was calculated using the Bazett formula (QT interval)/(RR interval)1/2(16). Significant arrhythmia complications were defined as 1) new or increased ventricular arrhythmias (new or increased ventricular ectopic beats or ventricular tachycardia, including torsade de pointes), as determined by the primary physicians caring for the patients; 2) significant bradycardia (heart rate ≤40 beats/min, pause ≥3.0 s, or resulting in pacemaker implantation or dose reduction); or 3) excessive QTc interval prolongation (>25% from baseline). Patients categorized as having no significant heart disease had no significant coronary or valvular disease or myocardial dysfunction. Patients with left ventricular hypertrophy by echocardiography or hypertension were not excluded from this category.
Statistical analysis was performed to determine univariate and multivariate baseline demographic, clinical, ECG and echocardiographic predictors of significant arrhythmia complications occurring before discharge from the hospital. Results are reported as mean value ± SD, unless otherwise stated. Continuous variables were compared using two-tailed Student ttests. Categoric variables were analyzed using chi-square or Fisher exact tests. Multiple logistic regression models, including stepwise methods, were used to measure associations between predictors and outcomes while adjusting for other variables (SPSS, Inc. and SAS). Differences were considered significant at p < 0.05.
One hundred twenty patients were admitted to the hospital for initiation of sotalol therapy for atrial arrhythmias (mean age 64 years; 64% male) (Table 1). Approximately 80% had underlying heart disease, and 20% had no heart disease. Of 117 patients (97.5%) with atrial fibrillation or atrial flutter, or both, 31 had both atrial fibrillation and flutter, and 11 had atrial fibrillation and other supraventricular arrhythmias, including atrial tachycardias. The three remaining patients had no history of atrial fibrillation and were treated for atrial tachycardia. Arrhythmias were paroxysmal in 84% of patients. The mean number of previously failed Vaughan Williams class I or III antiarrhythmic drugs was 1.93 (range 0 to 6). Twelve patients had a previous history of sustained ventricular tachycardia or ventricular fibrillation. Two patients had a history of proarrhythmia, both in response to class IC antiarrhythmic agents. None had a previous history of torsade de pointes. Permanent pacemakers had been implanted in 31 patients before initiation of sotalol therapy. The starting dose of sotalol was 192.0 ± 58.1 mg/day (median dose 160, range 80 to 320). The rhythm at initiation of sotalol was atrial fibrillation in 68 patients.
Significant arrhythmia complications
Significant arrhythmia complications were detected in 25 patients (20.8%) who experienced a total of 35 end points (Table 2). Seven patients (5.8%) developed new or increased ventricular arrhythmias, 8 (6.7%) developed excessively prolonged QT intervals, and 20 (16.7%) developed significant bradycardia. The clinical recognition of significant arrhythmia complications prompted a change in therapy during the hospital period in 21 patients (17.5%). Pacemakers were implanted in 3 patients (2.5%); sotalol was discontinued in 5 (4.2%); sotalol dosage was reduced in 18 (15.0%); and potassium was repleted in 1 (0.8%).
Prolonged QTc interval
Of the eight patients with a prolonged QTc interval, all except two had either concomitant bradycardia (three patients), new or increased ventricular arrhythmias (two patients) or bradycardia and new or increased ventricular arrhythmias (one patient). The remaining two patients had prolongation of the QTc interval to 606 and 656 ms, with the latter having bradycardia to 45 beats/min and hypotension that triggered discontinuation of sotalol. Sotalol dosage was reduced in four patients and discontinued in two. Potassium was repleted in one patient. The maximal QTc interval in the patients who developed a QTc interval >25% from baseline ranged from 544 to 656 ms.
Significant bradycardia occurred in 20 patients (16.7%) (Table 2). Fourteen patients (11.7%) had minimal heart rates ≤40 beats/min, and 5 (4.2%) had pauses ≥3.0 s. Six patients (5.0%) had minimal heart rates ≤35 beats/min. One patient had marked bradycardia requiring temporary pacing before electrical cardioversion. After cardioversion he developed complete atrioventricular block and subsequently underwent permanent pacemaker implantation. Three patients developed pauses >3.0 s after spontaneous cardioversion to sinus rhythm from atrial fibrillation, and three developed bradycardia to heart rates ≤40 beats/min after cardioversion. One patient had sinus arrest for 30 s after electrical cardioversion that was treated with atropine. Therapy changes were prompted in 18 patients (15.0%), with dosage reductions in 16 and permanent pacemaker implantation in 3 (2.5%).
New or increased ventricular arrhythmias
All seven patients with new or increased ventricular arrhythmias had nonsustained ventricular tachycardia that was new or increased after initiation of sotalol (Tables 2 and 3). ⇓Of these seven patients, two had nonsustained torsade de pointes that was asymptomatic and nocturnal in both. One patient, a 78-year old woman with normal left ventricular function, left ventricular hypertrophy and a pacemaker for complete heart block, had undergone aortic valve replacement. She developed ventricular bigeminy and 11-beat torsade de pointes occurring on day 6 after sotalol initiation. This event, which occurred during sotalol therapy of 160 mg twice daily and a pacemaker programmed to a lower rate limit of 60 beats/min, was associated with mild hypokalemia (K+3.4 mmol/liter) during diuretic therapy and resolved on repletion of potassium. The baseline QT interval was 320 ms (QTc 391 ms) and had increased to 484 ms (QTc 548 ms) at the time of the torsade de pointes. The second patient was a 74-year old man with coronary artery disease, moderate left ventricular dysfunction, ventricular tachycardia and an implantable cardioverter-defibrillator (ICD) with antibradycardic pacing programmed to VVI at 40 beats/min. He had received sotalol (160 mg twice daily) for 3 days, which was decreased to 120 mg twice daily after conversion of atrial fibrillation to sinus bradycardia during an ICD shock delivered for assessment of ventricular defibrillation safety margin. This patient developed bradycardia-associated nonsustained torsade de pointes on day 4 after sotalol initiation, the night after conversion to sinus rhythm, during VVI pacing at the lower rate limit, and was treated with a further reduction in sotalol dosage. His baseline QT and QTc intervals were 412 and 435 ms, respectively, increasing to 472 and 443 ms with sotalol, respectively. Three of seven cases of new or increased ventricular arrhythmias occurred in the early morning hours between midnight and 2:30 am. Two patients had an increase in ventricular ectopic frequency detected with monitoring that resulted in discontinuation of sotalol in one patient. The other patient underwent a reduction in dosage for increased ventricular ectopic beats, nonsustained ventricular tachycardia and concomitant significant bradycardia. All seven patients with new or increased ventricular arrhythmias underwent a change in treatment during the hospital period: Three patients were treated with reduction in dosage, three with discontinuation of sotalol, one with repletion of potassium and one with implantation of a permanent pacemaker.
Time to detection of significant arrhythmia complications
The mean time to the earliest detection of significant arrhythmia complications was 2.1 ± 2.5 days (range 0.5 to 12) after sotalol initiation (Fig. 1). The time to detection of new or increased ventricular arrhythmias was 3.2 ± 1.8 days (range 1 to 6), significant bradycardia 1.9 ± 1.8 days (range 0.5 to 8) and excessively prolonged QTc interval 4.4 ± 4.2 days (range 0.5 to 12). The time monitored in the hospital during sotalol therapy after its initiation was 5.9 ± 6.8 days (range 1 to 33) in the total group. Criteria for any significant arrhythmia complication, as defined in the present study, were met in 17 and 22 of the 25 patients with complications within 2 and 3 days of hospital admission, respectively. Except for the patient with torsade de pointes occurring in the setting of hypokalemia, the earliest significant arrhythmia complication detection was within 3 days in all patients with new or increased ventricular arrhythmias. Nocturnal (between 10 pmand 8 am) significant bradycardia was detected in 12 patients. Both episodes of torsade de pointes occurred nocturnally.
Significant arrhythmia complications in patients without underlying heart disease
Of the 24 patients without underlying heart disease, 3 (12.5%) developed significant arrhythmia complications. Two patients experienced bradycardia (minimal rates 37 and 39 beats/min, respectively), prompting a decrease in dosage in one patient. New or increased ventricular arrhythmias (nonsustained ventricular tachycardia and increased ventricular ectopic beats) were detected in one patient and led to discontinuation of sotalol.
Predictors of significant arrhythmia complications
Univariate and multivariate analyses showed that initial dosage was not a significant determinant of the development of significant complications (Table 4). In the total group, the mean starting dose of sotalol was 192 mg/day (median dose 160, range 80 to 320). There were no significant differences in initial or discharge doses between groups that did or did not develop significant arrhythmia complications, although there was a trend toward higher initial and lower discharge dosages in patients who had developed significant arrhythmia complications. Baseline ECG intervals, including QT, QTc, JT or corrected JT intervals, were not predictive of significant in-hospital arrhythmia complications. Concomitant digoxin or beta-adrenergic or calcium channel blocking agent (diltiazem or verapamil) use also did not predict complications. Age, creatinine clearance, ventricular function, previous history of ventricular arrhythmias and baseline rhythm also did not predict significant arrhythmia complications.
The absence of a permanent pacemaker before sotalol initiation was the only variable identified as a significant predictor of arrhythmia complications by the p < 0.05 significance criterion. The statistical significance of this variable was robust to adjustments for marginal predictors (p < 0.05 to p < 0.20). Similarly, adjustment for the absence of a pacemaker did not magnify the marginal effects of the other variables. These same findings were achieved using stepwise selection methods. Only 2 (6.5%) of 31 patients with a permanent pacemaker developed significant arrhythmia complications compared with 23 (25.8%) of 89 without a pacemaker (p = 0.022). One patient with a pacemaker developed torsade de pointes associated with hypokalemia, and another had excessive prolongation of the QTc interval.
No significant predictors for new or increased ventricular arrhythmias could be identified.
One hundred four patients were discharged from the hospital with sotalol therapy (mean dose at hospital discharge 227 ± 77 mg/day). Three of the 120 patients with initial sotalol therapy died after a mean follow-up period of 7.8 months. All died of noncardiac causes (massive cerebrovascular accident, pneumonia, esophageal cancer). No patients with significant arrhythmia complications that developed in the hospital died during follow-up. Pacemakers were implanted in four patients after discharge, one in a patient who had experienced significant in-hospital bradycardia and QTc interval prolongation. Of the 20 patients who experienced significant in-hospital arrhythmia complications and were discharged with sotalol, the drug was subsequently discontinued in 9, a pacemaker was implanted in 1 (who had experienced significant in-hospital bradycardia and QTc interval prolongation), and increased ectopic activity was noted in 1 (1 month later). Of the 84 patients discharged with sotalol who had not experienced significant in-hospital arrhythmia complications, pacemakers were implanted in 3, and new or increased ventricular arrhythmias were noted in 3 (2 with increased ectopic activity after initiation of sotalol, 2 with incessant or increased supraventricular tachycardia 2 months later). Sotalol was discontinued in 44 patients (1 because of QT interval prolongation, 2 because of new or increased ventricular arrhythmias and the remainder because of side effects or inefficacy).
The current study shows that the incidence of significant arrhythmia complications during initiation of sotalol therapy appears significant enough to warrant inpatient initiation. Bradycardia is the most frequent complication. Baseline patient characteristics or ECG variables could not be used to predict a low risk group, although patients with pacemakers were at lower risk for complications.
Previous studies of antiarrhythmic drug safety
Randomized studies of antiarrhythmic agents used for ventricular arrhythmias have demonstrated that the empiric use of class I antiarrhythmic drugs can be associated with the possibility of worsened survival (17–19). Although a potential exception is amiodarone, an antiarrhythmic agent with class III potassium channel blocking activity (20–23), the recent premature termination of the Survival With Oral d-Sotalol (SWORD) trial (24), a study for ventricular arrhythmias using the class III drug d-sotalol, highlights the need for continued caution, even with this group of antiarrhythmic agents.
Similar caution has been engendered by recent studies of antiarrhythmic drug use for atrial fibrillation. A meta-analysis of quinidine use for atrial arrhythmias (25)showed a threefold increase in overall mortality in patients treated with quinidine compared with placebo (2.9% vs. 0.8%), although the numbers of cardiovascular deaths were small. Although not randomized, antiarrhythmic drug use in the Stroke Prevention in Atrial Fibrillation (SPAF) trial (26)was associated with an increase in cardiac and arrhythmic death in patients with a history of congestive heart failure. These concerns have led to the initiation of a multicenter trial of ventricular rate control versus maintenance of sinus rhythm in patients with atrial fibrillation (the Atrial Fibrillation Follow-Up Investigation Of Rhythm Management [AFFIRM]).
Previous studies on sotalol
A pooled analysis of 1,288 patients enrolled in premarketing controlled trials using sotalol (7)showed that symptomatic bradyarrhythmias occurred in 111 patients, with 2.7% discontinuing because of bradycardia. New or increased ventricular arrhythmias occurred in 4.3%, with 2% classified as minor and 2% as severe (1.9% torsade de pointes). Other studies (1–8)have shown the risk of proarrhythmia from sotalol to be comparable to that of class IA antiarrhythmic drugs, with a reported range of 1% to 8%.
Sotalol causes a concentration-dependent lengthening of the QT interval (27,28). The drug increases action potential duration and refractory periods in cardiac tissue, most likely by a reduction in the delayed rectifier potassium current (IK), with selective blockade of the rapid component (IKr) (29–31). A small reduction in the inward rectifier potassium current (IK1) and blockade of the transient outward potassium current (Ito) in Purkinje fibers may also contribute to its action (32). Excessive prolongation of the QT interval by sotalol has been associated with the risk for significant arrhythmia proarrhythmia, particularly from torsade de pointes, in some studies (6,7), although this has not been uniformly reported (33).
Sotalol also has substantial beta-adrenergic blocking effects, with a potency similar to that of propranolol (34,35). The effects of sotalol on repolarization exhibit reverse use dependence, with a smaller effect at rapid rates and a more prominent effect at slow heart rates (14,15). This property may not only limit its effectiveness as an antiarrhythmic drug at rapid tachycardia rates, but could also produce more pronounced QT interval prolongation and risk for torsade de pointes during bradycardia. Bradyarrhythmias may not be readily or promptly detected during outpatient management because minimal bradycardia rates usually occur nocturnally during sleep.
However, because of the dose-dependent effects of sotalol, outpatient initiation has been performed in some patients, beginning with low doses. Nevertheless, torsade de pointes has been reported to occur early after initiation in patients treated for atrial fibrillation, after restoration of sinus rhythm or after an increase in dosage (2,7). Kehoe et al. (6)reported that proarrhythmia occurred in 31 (6.4%) of 481 patients treated for sustained ventricular arrhythmias. These events occurred within the acute loading phase of therapy in 23 patients (74%). Of the 15 patients who experienced torsade de pointes, all but 2 experienced it within 8 days of initiating or increasing sotalol dosage.
The current study
Incidence of complications
Approximately one in five patients initiated sotalol therapy appeared to benefit from inpatient monitoring and surveillance. Most complications consisted of significant bradyarrhythmias; however, asymptomatic nocturnal torsade de pointes was detected.
Criteria for significant arrhythmia complications
The criteria for serious complications in the present study were selected to detect significant arrhythmias or QTc prolongation that might require prompt intervention, including dosage adjustments, discontinuation of the drug or pacing. Such expeditious detection and intervention might be difficult to achieve during outpatient management. Potentially serious concomitant QTc interval prolongation and bradycardia did occur in a majority of the patients with new or increased ventricular arrhythmias, which may have facilitated earlier intervention. A lesser degree of bradycardia and QTc interval prolongation also occurred. Hypotension, congestive heart failure, bronchospasm or other potential side effects of sotalol occurred as well but were not considered immediately life threatening.
Symptoms were not considered necessary for inclusion as a serious complication because the primary purpose of the present study was to determine the need for in-hospital monitoring. Symptomatic arrhythmias, although potentially serious, would generally prompt patients to call or return for medical attention. However, asymptomatic arrhythmias may not prompt medical attention early enough to avoid serious arrhythmias in an outpatient, yet could still be a harbinger of impending serious complications. In the current study, half of the patients with significant bradycardia and both patients with torsade de pointes had this event recorded during nocturnal hours. Patients may also be at an especially high risk for proarrhythmia as inpatients or outpatients, despite being asymptomatic, if they pharmacologically convert to sinus rhythm on sotalol after long-standing atrial fibrillation (2)and have significant bradycardia afterward.
Duration of hospital stay required to detect complications
The time course of occurrence of significant arrhythmias is important in decisions regarding inpatient or outpatient initiation of antiarrhythmic drug therapy. Criteria for any significant arrhythmia complication as defined in the present study was met within 3 days in all but three patients, supporting a significant yield of hospital monitoring with a 3-day hospital period. All patients with new or increased ventricular arrhythmias met a criterion for significant arrhythmia complication within 3 days, except for the one patient with torsade de pointes who had hypokalemia.
Prediction of safety for outpatient initiation
Baseline ECG intervals, including heart rate and QT intervals, and sotalol dosage were not significantly predictive of the development of significant arrhythmia complications and therefore could not be used to determine in whom sotalol therapy might be safely started on an outpatient basis. These findings are in contrast to a previous study (7)in which patients who developed severe proarrhythmia had longer mean baseline QT and QTc intervals. Our findings are consistent with those of Kehoe et al. (6)who reported that proarrhythmia was not related to sotalol dose, baseline heart rate and QT intervals. Our study also showed that the paroxysmal or chronic nature of atrial fibrillation was not predictive of complications. In addition, the absence of heart disease did not identify a low risk group for significant arrhythmia complications, although torsade de pointes did not occur in this group. However, the small sample size of this subgroup limits generalization of this observation.
The only significant predictor of significant arrhythmia complications was the presence or absence of a permanent pacemaker. Pacemakers would be expected to prevent the significant bradyarrhythmias that could be induced by sotalol therapy. However, the two cases of torsade de pointes occurred in one patient with a pacemaker and one with an ICD. These cases indicate that pacing may not be entirely protective of proarrhythmia, particularly at low pacing rates. Patients who are not protected from bradycardia with a pacemaker may benefit the most from inpatient monitoring, but hospital admission may still be advisable for some, if not all, patients with pacemakers. In addition, careful avoidance of hypokalemia and programming of backup pacing rates to higher rates might be prudent if sotalol is being initiated.
Long-term follow-up did not reveal deaths from late proarrhythmia. This finding may reflect either the limitations from sample size or the effect of risk stratification and interventions, such as dosage reductions, drug discontinuation or pacing, that were prompted during inpatient monitoring or subsequent outpatient surveillance. Several patients did require pacemaker implantation after discharge, and two patients had subsequent proarrhythmia detected, although torsade de pointes did not occur.
The present study was a retrospective review, with significant arrhythmia complications detected by chart review. The criteria for significant complications were retrospectively defined. The determination of new or increased ventricular arrhythmias was dependent on clinician assessment rather than specific numeric criteria. The time to detection of significant arrhythmia complications most likely underestimates the time to clinical recognition of these complications, underscoring the need for adequate duration of monitoring and the need for prospective assessment of the yield of inpatient monitoring or safety of outpatient drug initiation if the latter is pursued. One potential difference in the manner in which sotalol therapy may be initiated in outpatients is in the starting dose or rate of dosage escalation. Outpatient management may include initiation at lower doses with slower dose titration. However, the median starting dose in the current inpatient study was 160 mg/day, and our conclusions do not appear to be affected by dose-dependent effects.
A significant incidence (20.8%) of significant arrhythmia complications occurred during sotalol loading that could be detected by in-hospital monitoring. These events triggered changes in therapy in 17.5% of patients with initiation of sotalol. The most frequent serious complication detected was significant bradycardia (16.7%). The incidence of new or increased ventricular arrhythmias was 5.8%, with a 1.77% incidence of torsade de pointes. When present, bradycardia and torsade de pointes occurred frequently during nocturnal hours. Most significant arrhythmia complications were detected within the first 3 days of monitoring. Baseline ECG intervals, including QT intervals, or rhythm were not predictive of these complications. Although the absence of overt heart disease did not identify a low risk subgroup for the significant arrhythmia complications of bradycardia and nonsustained ventricular tachycardia, torsade de pointes did not occur in this group. However, generalization of this observation should be limited by the small sample size of this subgroup. The absence of a permanent pacemaker before sotalol initiation was the only significant predictor of the development of significant arrhythmia complications.
Patients who are not protected from bradycardia with a pacemaker should be admitted to the hospital for inpatient monitoring during sotalol initiation. However, even in patients with a pacemaker, careful avoidance of hypokalemia, programming of backup pacing rates to higher rates and hospital admission for some, if not all, of these patients as well may be prudent if sotalol is being initiated. The yield of hospital monitoring after initiation of sotalol suggests that at least 72 h of observation may be beneficial.
We are grateful for the helpful reviews of Eric J. Topol, MD and Clifford V. Harding III, MD, PhD, as well as the secretarial assistance of Rosemarie Capone and Shelly White.
- electrocardiogram, electrocardiographic
- implantable cardioverter-defibrillator
- Received December 16, 1997.
- Revision received March 2, 1998.
- Accepted March 16, 1998.
- American College of Cardiology
- Antman E.M.,
- Beamer A.D.,
- Cantillon C.,
- McGowan N.,
- Friedman P.L.
- Kay G.N.,
- Plumb V.J.,
- Arciniegas J.G.,
- Henthorn R.W.,
- Waldo A.L.
- Wang J.,
- Bourne G.W.,
- Wang Z.,
- Villemaire C.,
- Talajic M.,
- Nattel S.
- Bazett H.
- Nademanee K.,
- Singh B.N.,
- Stevenson W.G.,
- Weiss J.N.
- Ceremuzynski L.,
- Kleczar E.,
- Krzeminska-Pakula M.,
- et al.
- Burkart F.,
- Pfisterer M.,
- Kiowski W.,
- Follath F.,
- Burckhardt D.
- Cairns J.A.,
- Connolly S.J.,
- Gent M.,
- Roberts R.
- Coplen S.E.,
- Antman E.M.,
- Berlin J.A.,
- Hewitt P.,
- Chalmers T.C.
- Flaker G.C.,
- Blackshear J.L.,
- McBride R.,
- Kronmal R.A.,
- Halperin J.L.,
- Hart R.G.
- Funck-Brentano C.
- Sanguinetti M.C.,
- Jurkiewicz N.K.
- Sanguinetti M.C.
- Carmeliet E.
- Lazzara R.