• 
• 

Three Part Question

In a patient with [atrial fibrillation requiring rate limitation], is [digoxin better than placebo] at achieving [control of ventricular rate]?

Clinical Scenario

A 57 year old woman attends the Emergency Department with newly diagnosed atrial fibrillation of uncertain duration. You decide to treat her by ventricular rate limitation and wonder whether you should use digoxin.

Search Strategy

Medline 1966 to 01/2005 using Ovid Interface.
Embase 1974 to 06/2005 using Dialog DataStar interface.
Medline:[(exp atrial fibrillation OR atrial fibrillation.mp) AND (exp digoxin.mp OR digoxin.mp OR exp digitalis OR exp digitalis glycosides OR digitalis.mp) AND (exp placebo OR exp placebo effect OR placebo.mp)]
Embase:[(atrial ADJ fibrillation) AND (digoxin OR digitalis OR glycoside$) AND (placebo)}.
The Cochrane library was also searched (accessed 02/2005).
The articles obtained had their references scrutinised for further articles.

Search Outcome

The Medline search produced a total of 81 articles, Embase 195, Cochrane revealed 8 papers, and 2 further studies were discovered by the reference review. A total of 8 studies were found to be relevant to the three part question and of sufficient quality.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Falk 1987	36 patients with newly diagnosed AF (<7/7 duration & HR > 85 /min) given digoxin capsules (1.4mg over 14 hours) or placebo.	Double-blind RCT	1. Cardioversion rate.	1. There was no significant difference in cardioversion rates between digoxin and placebo (50% digoxin vs 44% placebo)	1. Small study 2. Inadequate data given about the changes in heart rate
			2. HR changes	2. The authors claimed that there was a significant slowing of HR in the digoxin group but not in the placebo group but they gave no figures to back this up.
			Both over 18 hours
Rawles 1990	72 cardiology out-patients with evidence of paroxysmal AF (> 5 beats & < 24 hours) on ambulatory monitoring. 31 (43%) patients were taking digoxin (5 with a beta-blocker, 3 with a calcium channel blocker)	Cohort	1. Duration of episodes of AF	1. 17 patients had AF < 30 mins: 76% were on digoxin.	1. Small study 2. Not an ED population 3. Patients were chosen consecutively with no evidence of adequate matching of the groups. 4. No data was given regarding confounding drugs used in the non-digoxin group.
			2. Ventricular rate during the paroxysms.	2. There was no significant difference in HR during episodes of AF between those taking digoxin (mean 140/min) and those not (mean 134/min).
			Patients were continuously monitored for 48 hours.
Galun 1991	27 patients attending an ED with paroxysmal AF (>30 mins), 13 of whom were taking long-term digoxin and 14 were on no treatment.	Cohort	1. Ventricular rate while in AF	1. There was no significant difference in HR between the groups (118 +/- 16/min digoxin: 121 +/- 15/min control)	1. Small study 2. Consecutive patients were chosen in both groups but the data presented showed them to be well-matched 3. No evidence was given on frequency of episodes prior to attendance at ED.
			2. Digoxin levels	2. There was no correlation between digoxin level and ventricular response rate (r = 0.2)
Hnatkova 1996	44 cardiology out-patients with paroxysmal AF given digoxin or placebo.	Randomised double-blind, cross-over trial.	Mean & standard deviation RR interval during each episode of AF > 2 mins.	1. There was no significant difference in mean RR interval during digoxin. 2.The standard deviation of the lowest RR interval was significantly reduced on digoxin (p<0.0001)	1. Small study 2. Not an ED population 3. No details were given re dose and route of administration of digoxin given.
Jordaens 1997	39 cardiology patients with symptomatic AF (<1/52 duration) with a HR >100/min. They were given digoxin 1.25mg iv over 8 hours (in 3 boluses) or placebo.	RCT	1. Cardioversion rates @ 12 hours	1. There was no significant difference in cardioversion rates (47.4% digoxin cf 40% placebo)	1. Small study 2. Digoxin was given as rapid boluses (0.75mg over 10 mins then 0.25mg over 5 mins 3. The patients in the digoxin had longer duration of AF than placebo (49+/-43 hours cf 15+/-21 hours in placebo group).
			2. HR over the first 2 hours	2. Overall, there was a significant reduction in HR with digoxin (143 „³ 118/min p<0.002) but not in the placebo group (143 „³ 139/min)
The Digitalis in Acute Atrial Fibrillation (DAAF) Trial Group. 1997	239 patients with AF <1/52 duration, given iv digoxin (0.015-0.02mg/kg given in 3 divided doses over 6 hours) or placebo	Double-blind RCT	1. Cardioversion rate	1. There was no significant difference in cardioversion rates (46% placebo cf 51% digoxin [p=0.37])	Short duration of observation
			2. HR in patients who remained in AF	2. Digoxin was significantly better at slowing HR cf placebo (123 to 91/min digoxin; 120 to 116/min placebo [p=0.0001])
			Both observed over 16 hours
Murgatroyd 1999	80 cardiology out-patients with paroxysmal AF (≥1 episode/month) given digoxin 125-250µg (dose determined by nomogram based on renal function) The dose was adjusted according to digoxin levels (investigators & patients blind to results & 'dosage adjustments' were carried out for placebo). Treatment was continued until 2 episodes of AF separated by SR were documented, to a maximum of 61/7. After a 7/7 washout period, the patients were crossed over to the other limb of the study.	Double-blind, placebo-controlled cross-over trial.	1. Time to 2nd recorded episode of symptomatic AF.	1. There was a statistically significant increase in the time to the 2nd recurrences (13.5 days placebo, 18.7 days digoxin [p=0.037])	1. The study looked at symptomatic rather than total number of episodes. 2. Although this study is not concerned with the best treatment to terminate individual episodes of AF, it is of interest to consider agents that reduce frequency of episodes of paroxysmal AF when patients are being discharged post cardioversion.
			2. HR before & during paroxysms.	2. There was no significant difference
			3. Duration of paroxysms.	3. There was no significant difference
Atarashi 2002	94 medical out-patients with persistent AF (<1/52 duration) or paroxysmal AF (1-8 episodes/ month) given either digoxin (0.25mg) orally, or placebo. Patients with persistent AF were cardioverted (DC or class I anti-arrhythmics).	RCT	AF recurrence (¡Ý 1 min duration seen on ECG or symptomatic episode later confirmed on ECG) over a 6/12 follow-up period.	There was no significant difference in relapse rates between digoxin and placebo groups	Not an ED population

Comment(s)

AF <7/7 Duration Three studies1,5,6 looked at the effect of digoxin on atrial fibrillation of less than 7 days duration, and all found that digoxin was better than placebo at reducing ventricular response. The magnitude of this reduction was similar in two studies although baseline heart rates were different5,6. The third study did not report its data adequately1. The time taken to reach a significant reduction in heart rate was within 30 minutes in one study5 and in less than 2 hours in the other6. Of note, all three studies used different doses of digoxin (Falk 0.1mg/hour; Jordaens 0.16mg/hour; DAAF 0.23mg/hour), all of which were higher than those used in UK practice (the BNF recommends 1-1.5mg in divided doses over 24 hours = 0.0625mg/hour). All studies excluded patients with moderate to severe heart failure and were limited in their ability to detect complications due to digoxin therapy. In theory, the positive inotropic effect of digoxin may be of use for those patients with reduced or unknown left ventricular function. However, covariate analysis of the AFFIRM Study (The AFFIRM Investigators. Relationships between sinus rhythm, treatment, and survival in the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) Study. Circulation. 2004; 109; 1509-1513) indicated that long-term digoxin use (mean follow up in this study was more than 3 years) was associated with a higher mortality (for further discussion, see Rate vs Rhythm BET). Although such a study cannot entirely exclude the influence of confounding variables, it raises significant concerns about the safety of long term digoxin for patients with otherwise stable AF. Paroxysmal AF There is no evidence that digoxin has a statistically significant effect on heart rate in patients with paroxysmal atrial fibrillation2,3,4, and Rawles and colleagues2 even found indirect evidence that digoxin use may be associated with prolonged paroxysms. Clinically, this is probably not particularly important because the aim of treatment is to terminate the episode by cardioverting the patient, rather than rate-limiting them. There is no satisfactory evidence supporting its use for preventing recurrences of AF7,8. (PAF is associated with high sympathetic tone – digoxin works via parasympathetic system therefore unlikely to be effective)

Clinical Bottom Line

Digoxin is significantly better than placebo at controlling ventricular rate in atrial fibrillation of less than 7 days duration however the evidence for this is surprisingly limited. Although these studies excluded those patients with significant heart failure and were limited in their ability to detect complications, digoxin would seem to be a safe drug in the acute setting, if not with chronic use. There is no evidence for or against its use AF > 7/7 duration. Digoxin has no role in the treatment of paroxysmal AF and no anti-arrhythmic effect.

References

1. Falk RH, Knowlton AA, Bernard SA, Gotlieb NE, and Battinelli NJ. Digoxin for converting recent-onset atrial fibrillation to sinus rhythm: a randomized, double-blinded trial. Ann Intern Med. 1987; 106; 503-506.
2. Rawles JM, Metcalfe MJ, and Jennings K. Time of occurrence, duration, and ventricular rate of paroxysmal atrial fibrillation: the effect of digoxin. Br Heart J. 1990; 63; 225-227.
3. Galun E, Flugelman MY, Glickson M, and Eliakim M. Failure of long-term digitalization to prevent rapid ventricular response in patients with paroxysmal atrial fibrillation. Chest. 1991; 99; 1038-10
4. Hnatkova K, Murgatroyd FD, Camm AJ, and Malik M. Effect of digoxin on the ventricular rate during paroxysmal atrial fibrillation. PACE. 1996; 19; 1968-1971.
5. Jordaens L, Trouerbach J, Calle P, Tavernier R, Derycke E, Vertongen P, Bergez B, and Vanderkerckhove Y. Conversion of atrial fibrillation to sinus rhythm and rate control by digoxin in comparison to placebo. Eur Heart J. 1997; 18; 643-648.
6. The Digitalis in Acute Atrial Fibrillation (DAAF) Trial Group. Intravenous digoxin in acute atrial fibrillation: results of a randomized, placebo-controlled multicetre trial in 239 patients. Eur Heart J. 1997; 18; 649-654.
7. Murgatroyd FD, Gibson SM, Baiyan X, O'Nunian S, Poloniecki JD, Ward DE, Malik M, and Camm AJ. Double-blind, placebo-controlled trial of digoxin in symptomatic paroxysmal atrial fibrillation. Circulation 1999; 99; 2765-2770.
8. Atarashi H, Inoue H, Fukanami M, Sugi K, Hamada C, and Origasi H. Double-blind, placebo-controlled trial of apridine and digoxin for the prevention of symptomatic atrial fibrillation. Circulation Journal. 2002; 66; 553-556.