- Email: [email protected]
The Role of Holter Monitoring in Detecting Digitalis-Provoked Arrhythmias
The Role of Holter Monitoring in Detecting Digitalis-Provoked Arrhythmias· Carolyn Goren, M.D.,t and Pablo Denes,M.D.+
Certain arrhythmias detected on the electrocardiogram are considered to be reliable indicaton of digiCaUs intoxication. We have emnated the incidence of these arrhythmias on 24-hour electrocardiographic monitoring (Holter monitoring) in 69 consecutive patients who had serum levels of digoldn determined within 24 hoon of the onset of continuous electrocardiographic monitoring. According to the serum level of digoldn, the patients were divided into the fonowing three groups: (1) group 1 had 0 to 1.0 ng/m1 (31 patients); (2) group 2 had 1.1 to 2.0 ng/m1 (27 patients); and group 3 had ~ 2.1 ng/ml (11 patients). The fonowing arrhythmias were considered to reflect digitaUs-provoked arrhythmias: (1) penislent sinus bradycardia or sinus pauses (or both); (2) atrioventricular block; (3) paroxysmal atrial tachycardia with block; (4) accelerated junctional rhythm; (5) complex ventricular arrhythmias (multifocal ventric-
uIar premature beats, bigeminy and trigeminy, and pain); and (6) ventricular tachycardia. There was no significant difference in the incidence of these six categories of arrhythmias among the three groups. In addition, there was no slgulftcant difference in the mean semm level of digoldn for patients with and without the arrhythmia within each category. Ten of the 69 patients had combinations of three of the so-called dJgJtaIIs-provoked arrhythmias, with incidences among the three groups showing no significant differences. In conclnsJon, rhythms considered to be potentially due to digitalis intoxication are frequently observed in hospitalized patients undergoing 24-hour electrocardiographic monitoring, are frequently unrelated to the serum level of digoldn, and appear unlikely to reflect tme dJgJtaIis intoxication in many of these patients.
Manifestations of digitalis intoxication on the electrocardiogram 'include sinoatrial block, sinus arrest, atrioventricular block, atrial tachycardia with block, atrial fibrillation with high-grade atrioventricular block, nonparoxysmal atrioventricular junctional tachycardia, complex ventricular premature beats, and ventricular tachycardia.t" These disturbances in rhythm are encountered frequently during 24-hour electrocardiographic monitoring (Holter monitoring), often at times when a 12-lead ECG is unremarkable. Whether they should be considered to reflect digitalis intoxication is not known. To examine this question, a retrospective study was initiated in which all continuous electrocardiographic (Holter) monitoring recordings performed in this hospital over a 5~-month period were reviewed and the findings were related to available data on serum levels of digoxin.
(Autopak) of blood drawn within 24 hours of the onset of each monitoring was ascertained. Sixty-nine records (12 percent) were identified in which the basic rhythm was either normal sinus rhythm or atrial fibrillation, and for which concomitant data on serum levels of digoxin were available. Arrhythmias observed on 24-hour electrocardiographic monitoring were classified as to the presence or absence of the following six categories of disturbances in rhythm suspected of being digitalis-provoked anbythmias: (1) persistent sinus bradycardia or sinus pauses; (2) atrioventricular block (first, second, or third degree); (3) paroxysmal atrial tachycardia with block; (4) accelerated junctional tachycardia; (5) complex ventricular premature beats, eg, multifocal ventricular premature beats, bigeminy, trigeminy, pairs; and (6) ventricular tachycardia. Serum levels of digoxin were classified into the following three groups according to the probability of intoxication, from lowest to highest: (1) 0 to 1.0 ng/mI; (2) 1.1 to 2.0 ng/ml; and (3) ~2.1 ng/ml. Charts were reviewed to identify age, the serum level of potassium within 24 hours of the onset of continuous electrocardiographic monitoring, the stability of the dose of digoxin, cardiac-related drugs (antianbythmics, antihypertensives, diuretics, potassium supplements, and anticoagulants), and cardiac diagnosis.
MATERIALS AND METHODS
A total of 576 consecutive 24-hour electrocardiographic ( Holter) monitoring records obtained between Oct 1, 1978 and March 15, 1979 were reviewed. Availability of data on the serum level of digoxin obtained by radioimmunoassay °From the Section of Cardiology, Department of Medicine, Rush-Presbyterian St. Lukes Medical Center, Chicago. tFeIlow in Cardiology. tProfessor of Medicine. Reprint requests: Dr. Denes, 1725 West Harrison, Suite 114, Chicago 60612
CHEST, 79: 5, MAY, 1981
Analysis at Data Comparisons were made of the incidence of each disturbance in rhythm across digoxin groups by X2 tests, and of mean digoxin levels of patients with and without each disturbance in rhythm by Hests. In addition, since patients with digitalis intoxication may manifest more than one arrhythmia, patients with combinations of three of the six digitalis-provoked anbythmias were analyzed in similar fashion .
HOLTER MONITORING IN DIGITAUs-PROVOKED ARRHYTHMIAS 555
'11
Table l-Di~Pl'OI101uJdDUcar6aneea Cartlf.: Rlaytlam Paderw C,.,.ped 6,.. Senam l.-el of DipsIn *
'11
Digitali&-Provoked Arrhythmia
Group 1 (0-1.0 ng/ml)
Group 2 (1.1-2.0 ng/ml)
Group 3
Persistent sinus bradycardia or sinus pauses"
4/28 (14)
6/24 (25)
1/9 (11)
Atrioventricular block
10/31 (32)
7/27 (26)
4/11 (36)
Table S--l.eHh of "'--'II In PaIIe.... .wa and EIIe1a ~l'OI101uJd ..4rrlaytlamla*
Mean Serum Level or Digoxin, ng/ml (No. or Patients)
(~2.1
ng/ml)
1ft""*'
Digitali&-Provoked Arrhythmia
With Arrhythmia
.
Without Arrhythmia
Persistent sinus bradycardia or sinus pauses
1.44 ±0.54 (11)
1.29 ± 0.69 (SO)
Atrioventricular block
1.30 ±O.66 (21)
1.30±0.67 (48)
1.00 (1)
1.32 ±0.67 (60)
Paroxysmal atrial tachycardia with block--
1/28 (4)
0/24 (0)
0/9 (0)
Accelerated [unotional rhythm
1/31 (3)
5/27 (18)
1/11 (9)
Complex ventricular premature beats
Accelerated [unotional rhythm
1.64 ±0.56 (7)
1.30 ±0.68 (62)
21/31 (68)
17/27 (63)
8/11 (73)
Ventricular taehycardia
Complex ventricular premature beats
1.31 ±0.68 (46)
1.33±0.70 (23)
10/31 (32)
5/27 (18)
1/11 (9)
Ventricular tachycardia
1.12±0.58 (16)
1.39 ± 0.70 (53)
Combinations or 3 digitali&-provoked arrhythmias
1.27 ±0.54 (10)
1.97 ± 1.66 (59)
Combinations or 3 digitali&-provoked arrhythmias
5/31 (16)
4/27 (15)
1/11 (9)
No digitalis-provoked arrhythmias
7/31 (23)
6/27 (22)
2/11 (18)
-No significant differences among the three groups. Table values are numbers or patients; numbers within parentheses are percents. "Includes patients in normal sinus rhythm only.
REsuLTS The incidence of each disturbance in rhythm within each digoxin category is given in Table I, and mean levels of digoxin for each rhythm category are listed in Table 2. Persistent sinus bradycardia or sinus pauses were seen in 11 (18 percent) of 61 monitoring records in normal sinus rhythm and in 14 percent (4/28),25 percent (6/24). and 11 percent (1/9) of the three digoxin groups. respectively. Atrioventricular block occurred in 21 (30 percent) of all 69 electrocardiographic records and in 32 percent (10/31),26 percent (7/27), and 36 percent (4/11) of the three digoxin groups. Ninety-five percent of the atrioventricular block was first-degree block; there was one case of Wenkebach's type of second-degree atrioventricular block and another two cases of Mobitz type-2 atrioventricular block. Third-degree heart block was not observed in these records. Accelerated junctional tachycardia occurred in 7 (10 percent) of the 69 monitoring records and in 3 percent (1/21), 18 percent (5/27), and 9 percent (1/11) of the three digoxin groups. Complex ventricular premature beats occurred in 46 (frl percent) of all 69 records and in 68 percent (21/31), 63 percent (17/27), and 73 percent (8/11) of the three digoxin groups. Ventricular tachycardia was 556 GOREN, DENES
Paroxysmal atrial tachycardia with block
-No significant differences between levels.
seen in 16 (23 percent) of the 69 records and in 32 percent (10/31), 18 percent (5/27), and 9 percent ( 1/11) of the respective digoxin groups. Paroxysmal atrial tachycardia with block occurred in only one (2 percent) of the 61 patients, with a digoxin level less than 1 ng/ml. Ten (14 percent) of the 69 patients had combinations of three digitalis-provoked arrhythmias, with incidences across digoxin groups of 16 percent (5/31),15 percent (4/27), and 9 percent (1/11). Fifteen (22 percent) of the 69 patients had DO digitalis-provoked arrhythmias on monitoring, with 23 percent (7/31),22 percent (6/27), and 18 percent (2/11) incidences across the three digoxin groups. None of these differences in incidences were statistically significant. All patients were hospitalized at the time of 24hour electrocardiographic monitoring. Samples of blood for determining serum levels of digoxin were drawn at unspecified times after administration, but with stable dosage for all patients. The mean serum levels of digoxin for patients with and without each arrhythmia are listed in Table 2. The differences in the serum level of digoxin were not statistically significant. Information regarding patients' ages was available in 94 percent (65) of our cases, and analysis revealed no significant differences across the three digoxin groups (the mean age ± SD in group 1 was 68 ± 12 years, in group 2 was 69 + 18 years, and in group 3 was 76 ± 12 years). CHEST, 79: 5, MAY, 1981
Table 3--CCU'diGc Di~. in Pelle,.,. Crouped by Serum Lew!l 01 Di«o:Rn •
Diagnosis
DISCUSSION
No . of Group 1 Group 2 Group 3 Cases (n=31) (n=27) (n=11 )
Coronary arterial disease
26
12 (39)
11 (41)
3 (27)
Valvular heart disease
10
7 (23)
2 (7)
1 (9)
Hypertensive cardiovascular disease
3
3 (10)
0
0
Coronary and hypertensive heart disease
6
1 (3)
4 (15)
1 (9)
Coronary and valvular heart disease
3
1 (3)
0
2(18)
Valvular and hypertensive heart disease
2
1 (3)
0
1 (9)
Primary myocardial disease
5
2 (6)
1 (4)
2 (18)
0
1 (4)
0
4 (13)
8 (30)
1 (9)
Congenital heart disease Miscellaneous**
13
*No significant differences among groups. Table values are numbers of patients; numbers within parentheses are percents. "Includes no heart disease, chronic obstructive pulmonary disease, and sick sinus syndrome.
Serum levels of potassium were available in 74 percent of all 24-hour electrocardiographic monitoring records and did not differ significantly across digoxin groups (the mean serum level of potassium -+- SD in group 1 was 4.31 ± .51 mEq/L in group 2 was 4.37 -+- 0.35 mEq/L and in group 3 was 4.61 -+0.68 mEq/ L ). Patients in group 2 were found to be receiving significantly more cardiac-related drugs than were those in group 1 (P < 0.02), but there were no significant differences between the other two groups (the mean number drugs ± SD in group 1 was 1.50 -+- 1.04, in group 2 was 2.35 ± 1.52, and in group 3 was 1.36 -+- 1.29). Only four patients were receiving propranolol (one in group 1 and three in group 2) . Sinus bradycardia was seen in only one of these patients (group 2, serum level of digoxin of 1.9 ng/ ml). Cardiac diagnosis did not differ significantly across digoxin groups (Table 3). Twelve-lead ECGs recorded within 24 hours of onset of continuous monitoring record were available for a total of 51 (74 percent) out of 69 records. Five ECGs showed sinus bradycardia, nine showed first-degree atrioventricular block, and one revealed multifocal ventricular premature beats. No other disturbances in rhythm were detected on any of these ECGs.
CHEST, 79: 5, MAY, 1981
Digitalis intoxication is considered to be a clinical diagnosis based upon the presence of specific disturbances in cardiac rhythm, systemic symptoms, and elevated serum levels of digoxin. 1-4,8 The observed disturbances in rhythm relate to the effect of digitalis on the conduction and formation of cardiac impulses.v" The systemic symptoms include nausea, vomiting, anorexia, abdominal pain, gastrointestinal bleeding, agitation, psychosis, and disturbances of color vision.2,3,7,8 A high incidence of digitalis intoxication ranging from 10 to 35 percent has been reported in hospitalized patients receiving cardiac glycosides. 7,9-13 Nevertheless, manifestations of digitalis intoxication can also occur when the patient is not receiving digitalis. Neither the systemic symptoms nor the cardiac disturbances of rhythm and conduction are specific for digitalis intoxication; most may be found in other conditions. Although higher serum levels of digoxin correlate positively with clinical toxicity, the absolute value of a serum level of digoxin may be an unreliable indicator of digitalis intoxication.P'" The threshold for induction of digitalis intoxication may be variable from patient to patient and within a single patient on different occasions.? Furthermore, there are many factors that may affect cardiac sensitivity to preparations containing digitalis. These include advanced age ; sympathetic tone; thyroid status; serum electrolytic abnormalities; especially potassium, magnesium, and calcium; acute hypoxemia; and the nature and severity of the underlying heart disease. 3,4,7,8,13,14,20,21 We have evaluated the incidence of so-called "digitalis-provoked arrhythmias" during 24-hour electrocardiographic (Holter) monitoring in patients who had serum levels of digoxin determined within 24 hours of the onset of monitoring. The disturbances in rhythm that we evaluated were derived from published lists of rhythms considered compatible with digitalis intoxication.l'! Some, such as sinus bradycardia and first-degree atrioventricular block, are less widely accepted than others as manifestations of digitalis intoxication, but we have included them for the sake of completeness. In our sample the incidences of the disturbances in rhythm and conduction that we analyzed do not relate to the serum level of digoxin but are similarly distributed among those with low ( <1 ng/ml), accepted therapeutic (1.1 to 2 ng/ ml ), and elevated (~2.1 ng/ml) serum levels of digoxin. In addition, we found no significant difference in mean serum levels of digoxin when compared in patients with and without each of the various arrhythmias. We
HOLTER MONITORING IN DIGITALIS-PROVOKED ARRHYTHMIAS 557
also examined and found no significant differences in age, serum level of potassium, other concomitant cardiac medications, or underlying heart disease among the three groups of serum levels of digoxin. Furthermore, the ECGs of these patients, obtained at the time of 24-hour electrocardiographic (Holter) monitoring, failed to reveal the presence of digitalisprovoked arrhythmias in the majority of the cases. It appears that the rhythms suggestive of digitalis intoxication are nonspecific findings, frequently detected on long-term dynamic electrocardiographic monitoring in hospitalized patients with significant underlying heart disease, and may not correlate with serum levels of digoxin or with digitalis intoxication. A more specific test for digitalis intoxication is the reversibility of the arrhythmias and symptoms after the discontinuation of therapy with digitalis therapy.9,14,18,19 The retrospective nature of the present study did not permit observations regarding discontinuation of the drug with repeated 24-hour electrocardiographic (Holter) studies. The relatively low serum levels of digoxin in our sample, the lack of systemic symptoms attributable to digitalis intoxication, and the careful monitoring for the presence of arrhythmias attest to the current awareness of physicians of the potential for digitalis intoxication. The probability of digitalis intoxication in these hospitalized patients is indeed small. Rather than representing true digitalis intoxication, the arrhythmias detected on 24-hour electrocardiographic (Holter) monitoring in this sample probably reflect "background noise." Arrhythmias detected on such monitoring should raise the suspicion of digitalis intoxication but do not appear to be a reliable test for the diagnosis of digitalis intoxication unless coupled with other symptoms or findings, definite elevations of the serum level of digoxin, and resolution of arrhythmias after therapy with digitalis is withdrawn.
1 Wellens HJJ. The electrocardiogram in digitalis intoxication. Prog Cardioll976; 5:271-90. 2 Marcus FI, Ewy GA. Digitalis intoxication. In: Mason DT, ed, Cardiac Emergencies. Baltimore: Williams and Wilkins Co, 1976: 363-83.
558 GOREN, DENES
3 Kastor JA, Yurchak PM. Recognition of digitalis intoxication in the presence of atrial fibrillation. Ann Intern Med 1967; et, 1045-54. 4 Mason DT, Zelis R, Lee G, Huges J, Spann JF, Amsterdam EA. Current concepts and treatment of digitalis toxicity. Am J Cardiol 1971; 27:546-59. 5 Kastor JA. Digitalis intoxication in patients with atrial fibrillation. Circulation 1973; 47 :888-96. 6 Sodeman WA. Diagnosis and treatment of digitalis toxicity. N Engl J Med 1965; 273:35-37, 93·95. 7 Beller GA, Smith TW, Abelmann WH, Haber E, Hood WG Jr. Digitalis intoxication: a prospective clinical study with serum level correlation. N Engl J Med 1971; 284:989-97. 8 Moe GK, Farah AE. Digitalis and allied cardiac glycosides. In : Goodman, LS, Gilman A, ed. The pharmacologic basis of therapeutics. 5th ed. New York: Macmillan Publishing Co, Inc, 1975; 653-82. 9 Evered DC, Chapman C. Plasma digoxin concentrations and digoxin toxicity in hospital patients. Br Heart J 1971; 33:540-45. 10 Rodensky PL, Wasserman F. Observation on digitalis intoxication. Arch Intern Med 1961; 108:171-88. 11 Shapiro S, Slone D, Lewis GP, Jick H. The epidemiology of digoxin: a study in three hospitals. J Chronic Dis 1969; 22:361-71. 12 Hurwitz N, Wade OL. Intensive hospital monitoring of adverse reactions to drugs. Br Med J 1969; 1:531-36. 13 Smith TW. Digitalis toxicity: epidemiology and clinical use of serum concentration measurements. Am J Med 1975; 470-76. 14 Smith TW, Haber E. Digoxin intoxication : the relationship of clinical presentation to serum digoxin concentration. J Clin Invest 1970; 49:2377-86. 15 Shapiro W. Correlation studies of serum digitalis levels and the arrhythmias of digitalis intoxication. Am J Cardi011978; 41 :852-59. 16 Biddle TL, Weintraub M, Lasagna L. Relationship of serum and myocardial digoxin concentration to electrocardiographic estimation of digoxin intoxication. J Clin Pharmacol 1978; 18:10-15. 17 Schneider J, Ruiz-Torres A. Digitalis effect and blood concentration. Int J Clin Pharmacol 1977; 15:424-26. 18 Ingelfinger JA, Goldman P. The serum digitalis concentration: does it diagnose digitalis toxicity? N Engl J Med 1976; 294:8ff1-70.
19 Lasagna L. How useful are serum digitalis measurements? (editorial) . N Engl J Med 1976; 294:898-99. 20 Klein D, McInerny K, Levin PA, Ryan TJ. Digitalis tolerance in cardiomyopathy (abstract) . Am J CardioI1971; 39:271. 21 Butler VP Jr, Lindenbaum J. Serum digitalis measurements in the assessment of digitalis resistance and sensitivity. Am J Med 1975; 58:460-69.
CHEST, 79: 5, MAY, 1981
Certain arrhythmias detected on the electrocardiogram are considered to be reliable indicaton of digiCaUs intoxication. We have emnated the incidence of these arrhythmias on 24-hour electrocardiographic monitoring (Holter monitoring) in 69 consecutive patients who had serum levels of digoldn determined within 24 hoon of the onset of continuous electrocardiographic monitoring. According to the serum level of digoldn, the patients were divided into the fonowing three groups: (1) group 1 had 0 to 1.0 ng/m1 (31 patients); (2) group 2 had 1.1 to 2.0 ng/m1 (27 patients); and group 3 had ~ 2.1 ng/ml (11 patients). The fonowing arrhythmias were considered to reflect digitaUs-provoked arrhythmias: (1) penislent sinus bradycardia or sinus pauses (or both); (2) atrioventricular block; (3) paroxysmal atrial tachycardia with block; (4) accelerated junctional rhythm; (5) complex ventricular arrhythmias (multifocal ventric-
uIar premature beats, bigeminy and trigeminy, and pain); and (6) ventricular tachycardia. There was no significant difference in the incidence of these six categories of arrhythmias among the three groups. In addition, there was no slgulftcant difference in the mean semm level of digoldn for patients with and without the arrhythmia within each category. Ten of the 69 patients had combinations of three of the so-called dJgJtaIIs-provoked arrhythmias, with incidences among the three groups showing no significant differences. In conclnsJon, rhythms considered to be potentially due to digitalis intoxication are frequently observed in hospitalized patients undergoing 24-hour electrocardiographic monitoring, are frequently unrelated to the serum level of digoldn, and appear unlikely to reflect tme dJgJtaIis intoxication in many of these patients.
Manifestations of digitalis intoxication on the electrocardiogram 'include sinoatrial block, sinus arrest, atrioventricular block, atrial tachycardia with block, atrial fibrillation with high-grade atrioventricular block, nonparoxysmal atrioventricular junctional tachycardia, complex ventricular premature beats, and ventricular tachycardia.t" These disturbances in rhythm are encountered frequently during 24-hour electrocardiographic monitoring (Holter monitoring), often at times when a 12-lead ECG is unremarkable. Whether they should be considered to reflect digitalis intoxication is not known. To examine this question, a retrospective study was initiated in which all continuous electrocardiographic (Holter) monitoring recordings performed in this hospital over a 5~-month period were reviewed and the findings were related to available data on serum levels of digoxin.
(Autopak) of blood drawn within 24 hours of the onset of each monitoring was ascertained. Sixty-nine records (12 percent) were identified in which the basic rhythm was either normal sinus rhythm or atrial fibrillation, and for which concomitant data on serum levels of digoxin were available. Arrhythmias observed on 24-hour electrocardiographic monitoring were classified as to the presence or absence of the following six categories of disturbances in rhythm suspected of being digitalis-provoked anbythmias: (1) persistent sinus bradycardia or sinus pauses; (2) atrioventricular block (first, second, or third degree); (3) paroxysmal atrial tachycardia with block; (4) accelerated junctional tachycardia; (5) complex ventricular premature beats, eg, multifocal ventricular premature beats, bigeminy, trigeminy, pairs; and (6) ventricular tachycardia. Serum levels of digoxin were classified into the following three groups according to the probability of intoxication, from lowest to highest: (1) 0 to 1.0 ng/mI; (2) 1.1 to 2.0 ng/ml; and (3) ~2.1 ng/ml. Charts were reviewed to identify age, the serum level of potassium within 24 hours of the onset of continuous electrocardiographic monitoring, the stability of the dose of digoxin, cardiac-related drugs (antianbythmics, antihypertensives, diuretics, potassium supplements, and anticoagulants), and cardiac diagnosis.
MATERIALS AND METHODS
A total of 576 consecutive 24-hour electrocardiographic ( Holter) monitoring records obtained between Oct 1, 1978 and March 15, 1979 were reviewed. Availability of data on the serum level of digoxin obtained by radioimmunoassay °From the Section of Cardiology, Department of Medicine, Rush-Presbyterian St. Lukes Medical Center, Chicago. tFeIlow in Cardiology. tProfessor of Medicine. Reprint requests: Dr. Denes, 1725 West Harrison, Suite 114, Chicago 60612
CHEST, 79: 5, MAY, 1981
Analysis at Data Comparisons were made of the incidence of each disturbance in rhythm across digoxin groups by X2 tests, and of mean digoxin levels of patients with and without each disturbance in rhythm by Hests. In addition, since patients with digitalis intoxication may manifest more than one arrhythmia, patients with combinations of three of the six digitalis-provoked anbythmias were analyzed in similar fashion .
HOLTER MONITORING IN DIGITAUs-PROVOKED ARRHYTHMIAS 555
'11
Table l-Di~Pl'OI101uJdDUcar6aneea Cartlf.: Rlaytlam Paderw C,.,.ped 6,.. Senam l.-el of DipsIn *
'11
Digitali&-Provoked Arrhythmia
Group 1 (0-1.0 ng/ml)
Group 2 (1.1-2.0 ng/ml)
Group 3
Persistent sinus bradycardia or sinus pauses"
4/28 (14)
6/24 (25)
1/9 (11)
Atrioventricular block
10/31 (32)
7/27 (26)
4/11 (36)
Table S--l.eHh of "'--'II In PaIIe.... .wa and EIIe1a ~l'OI101uJd ..4rrlaytlamla*
Mean Serum Level or Digoxin, ng/ml (No. or Patients)
(~2.1
ng/ml)
1ft""*'
Digitali&-Provoked Arrhythmia
With Arrhythmia
.
Without Arrhythmia
Persistent sinus bradycardia or sinus pauses
1.44 ±0.54 (11)
1.29 ± 0.69 (SO)
Atrioventricular block
1.30 ±O.66 (21)
1.30±0.67 (48)
1.00 (1)
1.32 ±0.67 (60)
Paroxysmal atrial tachycardia with block--
1/28 (4)
0/24 (0)
0/9 (0)
Accelerated [unotional rhythm
1/31 (3)
5/27 (18)
1/11 (9)
Complex ventricular premature beats
Accelerated [unotional rhythm
1.64 ±0.56 (7)
1.30 ±0.68 (62)
21/31 (68)
17/27 (63)
8/11 (73)
Ventricular taehycardia
Complex ventricular premature beats
1.31 ±0.68 (46)
1.33±0.70 (23)
10/31 (32)
5/27 (18)
1/11 (9)
Ventricular tachycardia
1.12±0.58 (16)
1.39 ± 0.70 (53)
Combinations or 3 digitali&-provoked arrhythmias
1.27 ±0.54 (10)
1.97 ± 1.66 (59)
Combinations or 3 digitali&-provoked arrhythmias
5/31 (16)
4/27 (15)
1/11 (9)
No digitalis-provoked arrhythmias
7/31 (23)
6/27 (22)
2/11 (18)
-No significant differences among the three groups. Table values are numbers or patients; numbers within parentheses are percents. "Includes patients in normal sinus rhythm only.
REsuLTS The incidence of each disturbance in rhythm within each digoxin category is given in Table I, and mean levels of digoxin for each rhythm category are listed in Table 2. Persistent sinus bradycardia or sinus pauses were seen in 11 (18 percent) of 61 monitoring records in normal sinus rhythm and in 14 percent (4/28),25 percent (6/24). and 11 percent (1/9) of the three digoxin groups. respectively. Atrioventricular block occurred in 21 (30 percent) of all 69 electrocardiographic records and in 32 percent (10/31),26 percent (7/27), and 36 percent (4/11) of the three digoxin groups. Ninety-five percent of the atrioventricular block was first-degree block; there was one case of Wenkebach's type of second-degree atrioventricular block and another two cases of Mobitz type-2 atrioventricular block. Third-degree heart block was not observed in these records. Accelerated junctional tachycardia occurred in 7 (10 percent) of the 69 monitoring records and in 3 percent (1/21), 18 percent (5/27), and 9 percent (1/11) of the three digoxin groups. Complex ventricular premature beats occurred in 46 (frl percent) of all 69 records and in 68 percent (21/31), 63 percent (17/27), and 73 percent (8/11) of the three digoxin groups. Ventricular tachycardia was 556 GOREN, DENES
Paroxysmal atrial tachycardia with block
-No significant differences between levels.
seen in 16 (23 percent) of the 69 records and in 32 percent (10/31), 18 percent (5/27), and 9 percent ( 1/11) of the respective digoxin groups. Paroxysmal atrial tachycardia with block occurred in only one (2 percent) of the 61 patients, with a digoxin level less than 1 ng/ml. Ten (14 percent) of the 69 patients had combinations of three digitalis-provoked arrhythmias, with incidences across digoxin groups of 16 percent (5/31),15 percent (4/27), and 9 percent (1/11). Fifteen (22 percent) of the 69 patients had DO digitalis-provoked arrhythmias on monitoring, with 23 percent (7/31),22 percent (6/27), and 18 percent (2/11) incidences across the three digoxin groups. None of these differences in incidences were statistically significant. All patients were hospitalized at the time of 24hour electrocardiographic monitoring. Samples of blood for determining serum levels of digoxin were drawn at unspecified times after administration, but with stable dosage for all patients. The mean serum levels of digoxin for patients with and without each arrhythmia are listed in Table 2. The differences in the serum level of digoxin were not statistically significant. Information regarding patients' ages was available in 94 percent (65) of our cases, and analysis revealed no significant differences across the three digoxin groups (the mean age ± SD in group 1 was 68 ± 12 years, in group 2 was 69 + 18 years, and in group 3 was 76 ± 12 years). CHEST, 79: 5, MAY, 1981
Table 3--CCU'diGc Di~. in Pelle,.,. Crouped by Serum Lew!l 01 Di«o:Rn •
Diagnosis
DISCUSSION
No . of Group 1 Group 2 Group 3 Cases (n=31) (n=27) (n=11 )
Coronary arterial disease
26
12 (39)
11 (41)
3 (27)
Valvular heart disease
10
7 (23)
2 (7)
1 (9)
Hypertensive cardiovascular disease
3
3 (10)
0
0
Coronary and hypertensive heart disease
6
1 (3)
4 (15)
1 (9)
Coronary and valvular heart disease
3
1 (3)
0
2(18)
Valvular and hypertensive heart disease
2
1 (3)
0
1 (9)
Primary myocardial disease
5
2 (6)
1 (4)
2 (18)
0
1 (4)
0
4 (13)
8 (30)
1 (9)
Congenital heart disease Miscellaneous**
13
*No significant differences among groups. Table values are numbers of patients; numbers within parentheses are percents. "Includes no heart disease, chronic obstructive pulmonary disease, and sick sinus syndrome.
Serum levels of potassium were available in 74 percent of all 24-hour electrocardiographic monitoring records and did not differ significantly across digoxin groups (the mean serum level of potassium -+- SD in group 1 was 4.31 ± .51 mEq/L in group 2 was 4.37 -+- 0.35 mEq/L and in group 3 was 4.61 -+0.68 mEq/ L ). Patients in group 2 were found to be receiving significantly more cardiac-related drugs than were those in group 1 (P < 0.02), but there were no significant differences between the other two groups (the mean number drugs ± SD in group 1 was 1.50 -+- 1.04, in group 2 was 2.35 ± 1.52, and in group 3 was 1.36 -+- 1.29). Only four patients were receiving propranolol (one in group 1 and three in group 2) . Sinus bradycardia was seen in only one of these patients (group 2, serum level of digoxin of 1.9 ng/ ml). Cardiac diagnosis did not differ significantly across digoxin groups (Table 3). Twelve-lead ECGs recorded within 24 hours of onset of continuous monitoring record were available for a total of 51 (74 percent) out of 69 records. Five ECGs showed sinus bradycardia, nine showed first-degree atrioventricular block, and one revealed multifocal ventricular premature beats. No other disturbances in rhythm were detected on any of these ECGs.
CHEST, 79: 5, MAY, 1981
Digitalis intoxication is considered to be a clinical diagnosis based upon the presence of specific disturbances in cardiac rhythm, systemic symptoms, and elevated serum levels of digoxin. 1-4,8 The observed disturbances in rhythm relate to the effect of digitalis on the conduction and formation of cardiac impulses.v" The systemic symptoms include nausea, vomiting, anorexia, abdominal pain, gastrointestinal bleeding, agitation, psychosis, and disturbances of color vision.2,3,7,8 A high incidence of digitalis intoxication ranging from 10 to 35 percent has been reported in hospitalized patients receiving cardiac glycosides. 7,9-13 Nevertheless, manifestations of digitalis intoxication can also occur when the patient is not receiving digitalis. Neither the systemic symptoms nor the cardiac disturbances of rhythm and conduction are specific for digitalis intoxication; most may be found in other conditions. Although higher serum levels of digoxin correlate positively with clinical toxicity, the absolute value of a serum level of digoxin may be an unreliable indicator of digitalis intoxication.P'" The threshold for induction of digitalis intoxication may be variable from patient to patient and within a single patient on different occasions.? Furthermore, there are many factors that may affect cardiac sensitivity to preparations containing digitalis. These include advanced age ; sympathetic tone; thyroid status; serum electrolytic abnormalities; especially potassium, magnesium, and calcium; acute hypoxemia; and the nature and severity of the underlying heart disease. 3,4,7,8,13,14,20,21 We have evaluated the incidence of so-called "digitalis-provoked arrhythmias" during 24-hour electrocardiographic (Holter) monitoring in patients who had serum levels of digoxin determined within 24 hours of the onset of monitoring. The disturbances in rhythm that we evaluated were derived from published lists of rhythms considered compatible with digitalis intoxication.l'! Some, such as sinus bradycardia and first-degree atrioventricular block, are less widely accepted than others as manifestations of digitalis intoxication, but we have included them for the sake of completeness. In our sample the incidences of the disturbances in rhythm and conduction that we analyzed do not relate to the serum level of digoxin but are similarly distributed among those with low ( <1 ng/ml), accepted therapeutic (1.1 to 2 ng/ ml ), and elevated (~2.1 ng/ml) serum levels of digoxin. In addition, we found no significant difference in mean serum levels of digoxin when compared in patients with and without each of the various arrhythmias. We
HOLTER MONITORING IN DIGITALIS-PROVOKED ARRHYTHMIAS 557
also examined and found no significant differences in age, serum level of potassium, other concomitant cardiac medications, or underlying heart disease among the three groups of serum levels of digoxin. Furthermore, the ECGs of these patients, obtained at the time of 24-hour electrocardiographic (Holter) monitoring, failed to reveal the presence of digitalisprovoked arrhythmias in the majority of the cases. It appears that the rhythms suggestive of digitalis intoxication are nonspecific findings, frequently detected on long-term dynamic electrocardiographic monitoring in hospitalized patients with significant underlying heart disease, and may not correlate with serum levels of digoxin or with digitalis intoxication. A more specific test for digitalis intoxication is the reversibility of the arrhythmias and symptoms after the discontinuation of therapy with digitalis therapy.9,14,18,19 The retrospective nature of the present study did not permit observations regarding discontinuation of the drug with repeated 24-hour electrocardiographic (Holter) studies. The relatively low serum levels of digoxin in our sample, the lack of systemic symptoms attributable to digitalis intoxication, and the careful monitoring for the presence of arrhythmias attest to the current awareness of physicians of the potential for digitalis intoxication. The probability of digitalis intoxication in these hospitalized patients is indeed small. Rather than representing true digitalis intoxication, the arrhythmias detected on 24-hour electrocardiographic (Holter) monitoring in this sample probably reflect "background noise." Arrhythmias detected on such monitoring should raise the suspicion of digitalis intoxication but do not appear to be a reliable test for the diagnosis of digitalis intoxication unless coupled with other symptoms or findings, definite elevations of the serum level of digoxin, and resolution of arrhythmias after therapy with digitalis is withdrawn.
1 Wellens HJJ. The electrocardiogram in digitalis intoxication. Prog Cardioll976; 5:271-90. 2 Marcus FI, Ewy GA. Digitalis intoxication. In: Mason DT, ed, Cardiac Emergencies. Baltimore: Williams and Wilkins Co, 1976: 363-83.
558 GOREN, DENES
3 Kastor JA, Yurchak PM. Recognition of digitalis intoxication in the presence of atrial fibrillation. Ann Intern Med 1967; et, 1045-54. 4 Mason DT, Zelis R, Lee G, Huges J, Spann JF, Amsterdam EA. Current concepts and treatment of digitalis toxicity. Am J Cardiol 1971; 27:546-59. 5 Kastor JA. Digitalis intoxication in patients with atrial fibrillation. Circulation 1973; 47 :888-96. 6 Sodeman WA. Diagnosis and treatment of digitalis toxicity. N Engl J Med 1965; 273:35-37, 93·95. 7 Beller GA, Smith TW, Abelmann WH, Haber E, Hood WG Jr. Digitalis intoxication: a prospective clinical study with serum level correlation. N Engl J Med 1971; 284:989-97. 8 Moe GK, Farah AE. Digitalis and allied cardiac glycosides. In : Goodman, LS, Gilman A, ed. The pharmacologic basis of therapeutics. 5th ed. New York: Macmillan Publishing Co, Inc, 1975; 653-82. 9 Evered DC, Chapman C. Plasma digoxin concentrations and digoxin toxicity in hospital patients. Br Heart J 1971; 33:540-45. 10 Rodensky PL, Wasserman F. Observation on digitalis intoxication. Arch Intern Med 1961; 108:171-88. 11 Shapiro S, Slone D, Lewis GP, Jick H. The epidemiology of digoxin: a study in three hospitals. J Chronic Dis 1969; 22:361-71. 12 Hurwitz N, Wade OL. Intensive hospital monitoring of adverse reactions to drugs. Br Med J 1969; 1:531-36. 13 Smith TW. Digitalis toxicity: epidemiology and clinical use of serum concentration measurements. Am J Med 1975; 470-76. 14 Smith TW, Haber E. Digoxin intoxication : the relationship of clinical presentation to serum digoxin concentration. J Clin Invest 1970; 49:2377-86. 15 Shapiro W. Correlation studies of serum digitalis levels and the arrhythmias of digitalis intoxication. Am J Cardi011978; 41 :852-59. 16 Biddle TL, Weintraub M, Lasagna L. Relationship of serum and myocardial digoxin concentration to electrocardiographic estimation of digoxin intoxication. J Clin Pharmacol 1978; 18:10-15. 17 Schneider J, Ruiz-Torres A. Digitalis effect and blood concentration. Int J Clin Pharmacol 1977; 15:424-26. 18 Ingelfinger JA, Goldman P. The serum digitalis concentration: does it diagnose digitalis toxicity? N Engl J Med 1976; 294:8ff1-70.
19 Lasagna L. How useful are serum digitalis measurements? (editorial) . N Engl J Med 1976; 294:898-99. 20 Klein D, McInerny K, Levin PA, Ryan TJ. Digitalis tolerance in cardiomyopathy (abstract) . Am J CardioI1971; 39:271. 21 Butler VP Jr, Lindenbaum J. Serum digitalis measurements in the assessment of digitalis resistance and sensitivity. Am J Med 1975; 58:460-69.
CHEST, 79: 5, MAY, 1981