- Email: [email protected]
Residual atrial fibrillation and clinical consequences following postoperative supraventricular arrhythmias
International Journal of Cardiology 74 (2000) 125–132 www.elsevier.com / locate / ijcard
Residual atrial fibrillation and clinical consequences following postoperative supraventricular arrhythmias Mahmoud Loubani FRCSI, Mark St. J. Hickey FRCSI, Tom J. Spyt FRCS, ˜ Manuel Galinanes FRCS* Division of Cardiac Surgery /Department of Surgery, University of Leicester, Glenfield Hospital, Groby Road, Leicester, LE3 9 QP, UK Received 18 October 1999; received in revised form 15 December 1999; accepted 28 January 2000
Abstract Aims: This retrospective study investigated whether the supraventricular arrhythmias (SVA) observed during cardiac surgery are limited to or persist beyond the postoperative period, their clinical consequences and whether they are influenced by preoperative and postoperative factors. Methods: A total of 375 patients undergoing elective bypass graft surgery over a 15-month period by three surgeons were included. All patients had their preoperative medications continued to the day of surgery and prophylactic anti-arrhythmic medications were not used in any of the cases. Standard anaesthetic techniques were used. Rhythm disturbances were diagnosed by ECG. The arrhythmias were treated medically or by cardioversion. All patients were followed up for 6 months. Results: Postoperative SVA occurred in 25% of patients. The commonest arrhythmia was atrial fibrillation (89.4%), followed by atrial flutter (6.4%) and supraventricular tachycardia (4.2%). In 89.8% of the cases, the arrhythmias occurred within the first four postoperative days with a maximum incidence on the second day (27.7%). Atrial fibrillation was still present in 50% of patients at hospital discharge and in 39% at 6 months follow up. Patients with arrhythmias had a prolonged hospital stay (7.762.6 vs. 6.062.6 days; P,0.05). There was no hospital mortality in the study and the incidence of postoperative stroke was equal in the sinus rhythm and arrhythmia patients (1.1%). SVA were more frequent when cardioplegia was used to protect the heart (32%) than with intermittent ischaemia (9%; P,0.001). At 6 months follow up, the patients receiving cardioplegia also had a higher prevalence of atrial fibrillation than those operated with intermittent ischaemia (41% vs. 22%; P,0.05). The incidence of SVA and persistence of atrial fibrillation was unrelated to other preoperative and intraoperative factors. Conclusion: Postoperative supraventricular arrhythmias have a long-lasting effect on cardiac rhythm: patients with SVA have a high probability of remaining in atrial fibrillation at hospital discharge and 6 months after surgery. The occurrence of atrial fibrillation seems to be influenced by the type of myocardial protection used but this does not appear to exert harmful effects. 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Atrial fibrillation; Cardiac surgery; Follow up
1. Introduction Supraventricular arrhythmias (SVA) are the most common complication following cardiac surgery; their incidence has been reported to vary from 10 to *Corresponding author. Tel.: 144-116-250-2449; fax: 144-116-2321282. ˜ E-mail address: [email protected] (M. Galinanes).
50% [1], and it is known that they are associated with increased morbidity and length of hospitalization [2], all adding to the cost of surgery. However, the prevalence and clinical consequences of these arrhythmias once the patient has been discharged from hospital are unknown. Furthermore, the cause of these rhythm disturbances still remains unclear, although a number of factors including age [1], anatomical distribution of coronary artery disease [3],
0167-5273 / 00 / $ – see front matter 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S0167-5273( 00 )00229-1
126
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
renal failure [1], poor left ventricular function [4], abrupt withdrawal of b blockers [5], history of rheumatic heart disease [2], chronic obstructive airway disease [2] and increasing aortic cross clamp time [2,6] have been reported to affect the incidence of SVA. The aims of the present study were to investigate whether postoperative SVA are short-lived and confined to the early postoperative period or persistent beyond hospital discharge, their clinical consequences, and whether their incidence is influenced by preoperative or intraoperative factors.
2. Materials and methods
2.1. Patients and exclusion criteria This was a retrospective study in which all patients (n5375) undergoing elective coronary artery bypass grafting surgery for a period of 15 months in a single institution and in normal sinus rhythm preoperatively were included. Data were collected when all the patients had a follow-up of at least 6 months. Patients operated on as emergency or undergoing a concomitant surgical procedure and those with preoperative history of SVA were excluded from the study. Patients were operated on by three surgeons each using only one of the three following methods of myocardial protection: cold crystalloid cardioplegia (n5122), normothermic blood cardioplegia (n5147) and intermittent ischaemia without cardioplegia (n5 106).
2.2. Medication and anaesthesia All preoperative medications, including b blockers and calcium antagonist were continued up to the day of surgery, except acetylsalicylic acid, which was discontinued at least one week before surgery. Patients received standard pre-medication and anaesthesia and no anti-arrhythmic medication (i.e., digoxin, amiodrone, b blockers) was administered prophylacticaly during or after surgery.
2.3. Surgery Cardiopulmonary bypass was established with a two-stage venous cannula in the right atrium and arterial return to the ascending aorta. In patients operated on with cold crystalloid cardioplegia, the body was cooled to 288C. The aorta was then crossclamped and cold (4–88C) crystalloid cardioplegia (St. Thomas’ cardioplegic solution no. 1) infused into the aortic root every 20 min. Once the distal anastomoses were completed, the aortic-cross clamp was removed, rewarming was initiated and the proximal anastomoses were completed. In patients with warm blood cardioplegia, body temperature was maintained at normothermia. The heart was then arrested with warm blood cardioplegia (KCl 20 mmol / l) infused into the aortic root and after completion of each distal anastomosis, additional doses of cardioplegia (KCl 10 mmol / l for the second to fourth dose and 7 mmol / l for all further doses) were given. The proximal anastomoses were carried out after releasing the aortic cross-clamp. In patients protected with intermittent ischaemia, the body was cooled to 328C and, in contrast with the other patients, no cardioplegia was used. Instead, the distal anastomoses were performed during brief periods of ischaemia achieved by cross-clamping the ascending aorta. The ischaemic episodes were interspaced by a period of reperfusion that ranged from 3 to 10 min, during which time the proximal anastomoses of the grafts to the ascending aorta were performed.
2.4. Detection of arrhythmia SVA was taken to include atrial fibrillation, atrial flutter and supraventricular tachycardia. Patients were monitored by continuous bedside electrocardiogram (ECG) monitoring of lead II for the first 24 h after surgery. The occurrence of SVA was confirmed by rhythm strip. Beyond this period the arrhythmias were detected by regular four-hourly clinical assessment or earlier if any symptoms were reported and confirmed by ECG. Peri-operative myocardial infarction was diagnosed with ECG changes and measurement of serum creatinine kinase (MB fraction). All patients were reviewed at 6 weeks and 6 months postoperatively in an outpatient clinic and questioned
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
with regard to residual angina. An ECG was done on each visit to evaluate the cardiac rhythm.
2.5. Treatment of arrhythmias Following the diagnosis of SVA, patients received one or a combination of the following therapies: digoxin, amiodarone or b blockers. If the medical therapy resulted in re-establishing sinus rhythm or in control of the ventricular rate and the patient was asymptomatic then the medication was continued for 6 weeks. At the end of this period and coinciding with the first postoperative visit at the outpatient clinic, if the rhythm had reverted to sinus then the medication was discontinued, if still in atrial fibrillation other medications were added or cardioversion was considered. Symptomatic patients or failure of anti-arrhythmic medications to control the ventricular rate led to cardioversion at any time during the follow-up. The procedure for the treatment of SVA followed by the three operating surgeons was identical. The persistence of the arrhythmia in spite of treatment was not a reason to maintain patients in hospital if ventricular heart rate was between 60 and 100 beats per min and this was not associated with hemodynamic impairment.
2.6. Expression of results and statistical analysis Data were expressed as means6standard deviation and the comparisons between the groups were made by the ANOVA analysis and x 2 test. Univariate and multivariate analyses were carried out to assess the significance of preoperative and intraoperative factors in the development of arrhythmias. A value of P, 0.05 was taken to be statistically significant.
3. Results
3.1. Postoperative supraventricular arrhythmias Table 1 shows that up to a quarter of patients undergoing elective bypass graft surgery have SVA in the postoperative period. It also shows that the clinical characteristics were comparable in patients with sinus rhythm and in those exhibiting SVA. Thus,
127
factors such as age, sex, angina class and the presence of hypertension, diabetes, hypercholesterolaemia or renal impairment have no significant effect. Furthermore, in the present study the distribution of coronary artery disease, the state of the left ventricular function, the ischaemic time (i.e., mean aortic cross-clamp time), incidence of perioperative myocardial infarction and persistence of angina at 6 months follow-up do not appear to influence the occurrence of postoperative SVA. However, Table 2 shows that the type of myocardial protection used during surgery was a significant factor in determining the occurrence of postoperative SVA. Patients receiving cold crystalloid cardioplegia and warm blood cardioplegia had a comparable incidence of SVA (.30%) that was higher than in patients operated on with the use of intermittent ischaemia (,10%). In all instances, atrial fibrillation was the commonest arrhythmia and accounted for almost 90% of SVA. Fig. 1 demonstrates that most of the arrhythmias occur within the first 6 days of the postoperative period with the majority of them (73.4%) being initiated within the first 3 days. Only two patients exhibited atrial fibrillation after this period; one required readmission for 3 days and the other was still an in-patient as a result of a wound infection. A total of seven patients were cardioverted in the immediate postoperative period and a further 17 in the ensuing 6 months. However, the choice of treatment of the rhythm disturbance whether it was one or more pharmacological agents or a combination with cardioversion did not seem to have any influence on the persistence of atrial fibrillation in our study patients.
3.2. Clinical consequences Patients with SVA had a significantly prolonged postoperative hospital stay (7.762.6 days) than patients who remained in sinus rhythm (6.062.6 days; P,0.001). This was due to the time required for the control of the ventricular rate with medical treatment or cardioversion. It should be noted, however, that maintenance of atrial fibrillation with controlled heart rate of less than 100 beats per min was not a reason for delaying hospital discharge. Interestingly, as seen in Table 3 only 50% of patients presenting SVA were
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
128
Table 1 Analysis of the clinical characteristics and operative data as risk factors for the occurrence of postoperative supraventricular arrhythmias a Supraventricular arrhythmia
Sinus rhythm
Total number of patients Age (years) Sex (male:female)
94 (25.5%) 63.568.2 76:15
281 (74.5%) 61.969.0 225:56
NS NS
NYHA angina class 1 and 2 3 and 4 Hypertension Diabetes Hypercholesterolaemia Renal impairment (Cr.120 mmol / l)
32 62 30 12 40 5
(34.0%) (64.9%) (33.0%) (13.2%) (42.5%) (5.3%)
78 (27.7%) 203 (72.4%) 106 (37.3%) 51 (18.0%) 128 (45.5%) 13 (4.6%)
NS NS NS NS NS NS
Coronary artery disease Single vessel disease Double vessel disease Triple vessel disease RCA disease LMS disease
3 23 68 63 17
(3.1%) (24.4%) (72.3%) (62.3%) (18.0%)
12 (4,2%) 64 (22.7%) 205 (72.9%) 246 (87.4%) 65 (23.1%)
NS NS NS NS NS
13 (4.6%) 41.6615.6
NS NS
12 (4.2%) 82 (29.1%) 125 (44.4%) 52 (18.5%) 9 (3.2%) 1 (0.4%)
NS NS NS NS NS NS
Poor LV (EF,30%) Mean aortic cross clamp time (mm)
3 (3.2%) 41.9614.2
Number of grafts 1 2 3 4 5 6
5 24 45 18 2 0
Peri-operative MI Residual angina at 6 months post-operative a
(5.3%) (25.5%) (47.8%) (19.1%) (19.1%)
1 (1.1%) 2 (2.1%)
4 (1.4%) 5 (1.7%)
P value
NS NS
RCA, right coronary artery; LMS, left main stem; LV, left ventricle; EF, ejection fraction; Cr, creatinine, MI, myocardial infarction.
in sinus rhythm at the time of hospital discharge. Moreover, in spite of antiarrhythmic treatment, 48% of the patients still remained in atrial fibrillation at 6 weeks follow up, a percentage that was slightly reduced to 39% after 6 months. Table 3 also shows that the percentage of patients in atrial fibrillation was
significantly higher in the cardioplegic groups than in the intermittent ischaemia group. Thus, atrial fibrillation was still present in between two-thirds and a half of patients protected with cardioplegia, whereas only one fifth of those operated on using intermittent ischaemia were in atrial fibrillation. It is worth noting
Table 2 Influence of the type of myocardial protection on postoperative supraventricular arrhythmias
Atrial fibrillation Atrial flutter Supraventricular tachycardia Overall incidence a b
Cold crystalloid cardioplegia
Warm blood cardioplegia
Intermittent ischaemia
Total
34 (27.8%) 3 (2.4%)
41 (27.9%) 3 (2.0%)
9 (8.5%)a 0b
84 (89.4%) 6 (6.4%)
3 (2.4%) 40 (32.8%)
1 (0.6%) 45 (30.6%)
0b 9 (8.5%)a
4 (4.2%) 94 (25.5%)
P,0.001 vs. both cardioplegia groups. P,0.05 vs. both cardioplegia groups.
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
129
Fig. 1. Day of onset of supraventricular arrhythmia (SVA) following cardiac surgery.
that although the incidence of atrial fibrillation was similar in patients protected with cold crystalloid cardioplegia and those receiving warm blood cardioplegia at the time of hospital discharge, a significantly lower percentage of individuals remained in atrial
fibrillation in the latter group than in the former at 6 weeks and 6 months of follow-up. In considering the high prevalence of atrial fibrillation even after 6 months since operation, it is worth noting that with the exception of one patient developing a stroke in
130
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
Table 3 Persistence of atrial fibrillation a Atrial fibrillation
Cold crystalloid cardioplegia
Warm blood cardioplegia
Intermittent ischaemia
At hospital discharge At 6 weeks follow up At 6 months follow up Overall incidence
21 25 20 40
25 18 15 45
1 2 2 9
(52.5%) (62.5%) (50.0%) (32.8%)
(55.5%) (40.0%)c (33.3%)c (30.6%)
(1.1%)b (22.2)b (22.2%)b (8.5%)b
Total 47 (50%) 45 (47.8%) 4 (4.2%) 94 (25.5%)
a
The figures in brackets express the percentage of patients with atrial fibrillation of those that exhibited postoperative supraventricular arrhythmias. P,0.001 vs. both cardioplegia groups. c P,0.05 vs. cold crystalloid cardioplegia. b
the third postoperative day there were no further complications associated with arrhythmias. The incidence of stroke in patients with sinus rhythm was equal to that of patients with SVA (1.1%).
4. Discussion The present study has shown for the first time that more than a third of patients exhibiting SVA following bypass graft surgery still remain in atrial fibrillation after 6 months follow up and that this occurs in spite of the continuation of anti-arrhythmic medication. In addition, our study provides convincing evidence that the type of myocardial protection determines not only the incidence of postoperative SVA but also the prevalence of atrial fibrillation within the first 6 months after surgery. These results raise a number of important points that warrant further discussion.
4.1. Postoperative supraventricular arrhythmias SVA is the most common complication following bypass graft surgery and this is supported by our study. This shows that the overall incidence of SVA is 25% of elective graft surgery patients that did not have a previous history of rhythm disturbances, which compares with other studies reporting an incidence oscillating between 10 and 50% [1]. In this selected group of patients, SVA lengthened hospitalization but it did not increase morbidity. This contrasts with other studies including patients with graft and valve surgery alone and in combination, and patients with heart transplantation [2] in which SVA were associated to an increased incidence of stroke, increased occurrence of ventricular arrhythmias and a
greater need for implantation of pacemakers. In a recent study, Solomon et al. [7] have reported that no thromboembolic events occurred after cardiac surgery in 12 patients that were discharged in atrial fibrillation and were followed up for over 6 months. However, in their study all patients received warfarin anti-coagulation. In the present study, only one patient with SVA had a stroke in the third postoperative day and this was possibly attributed to the cardiac arrhythmia since other potential cardiac and vascular causes were ruled out. The distinction of the study population is important because SVA in patients with other types of surgery involving the implantation of intravascular foreign materials and with increased cardiac cavities dimensions and altered flow patterns and hemodynamics may be more predisposed to cardiovascular complications than patients undergoing graft surgery alone.
4.2. Residual atrial fibrillation The most striking finding of this study is that half of the patients experiencing postoperative SVA leave the hospital in atrial fibrillation and that a high proportion of these remain with the arrhythmia after 6 weeks and 6 months follow up. Furthermore, patients that do not develop SVA during the early postoperative period are unlikely to manifest with atrial fibrillation during the ensuing follow up. The high prevalence of atrial fibrillation after hospital discharge in patients with postoperative SVA may have important clinical implications since individuals in atrial fibrillation are at high risk of stroke [8] and death [9–11]. The absence of morbidity and mortality attributable to atrial fibrillation after hospital discharge in our study may be related to the insufficient number of patients entered in the study
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
and to the short follow up period. Another possibility is that all patients received warfarin and / or aspirin (150 mg / day) for the entire study period and this per se may have reduced the risk of stroke. The thesis that anti-platelet or anti-coagulation therapy may explain the results of the present study are supported by placebo-controlled trials in which warfarin was shown to substantially reduce stroke rates [12]. Clearly, further studies are required to define the clinical repercussions of the persistence of atrial fibrillation after cardiac surgery and to determine the best therapeutic strategies for its management.
4.3. Possible causes Our study demonstrates that both the incidence of postoperative SVA and the residual atrial fibrillation beyond the postoperative period depend on the type of myocardial protection used. Thus, the occurrence of arrhythmia was three times higher in patients protected with both blood and crystalloid cardioplegia than in those operated with short intermittent periods of ischaemia, a method that mimics the potent protective protocols of ischaemic preconditioning [13]. These results provide support to the thesis that ischaemic injury during cardiac surgery may be responsible for the development of postoperative arrhythmia and at first sight, one may be tempted to conclude that intermittent ischaemia is more effective than either blood or crystalloid cardioplegia. However, it may be possible that atrial and ventricular myocardium are differently protected during cardiac surgery. In this connection, it has been reported that atrial myocardium is less effectively cooled than ventricular myocardium when hearts are infused with hypothermic cardioplegia [14]. If this is the case, then it may be expected to have a greater degree of ischaemic injury in the atrium than in the ventricle and therefore an increased susceptibility to the development of postoperative SVA when cardioplegia is the protective modality used. In a previous study, Butler et al. [15] showed that intermittent ischaemia reduced the incidence of postoperative SVA from 25.6% in patients protected with hypothermic crystalloid cardioplegia to 19.3%; however the difference did not achieve statistical significance. In this study [15], a two-stage venous cannula was used to establish cardiopulmonary bypass in the cardioplegia
131
group whereas a basket cannula located in the right atrium was utilised in the intermittent ischaemia group, and this may have accounted for the higher incidence of SVA in patients treated with intermittent ischaemia in their study. Certainly, the use of a basket cannula will allow a major contact of blood with the atrium and probably a greater distension of the cavity that may increase the ischaemic injury of the atrium. This possibility is supported by the findings of Rousou et al. [16] that the use of a single cannula is associated with less effective cooling of the right atrium and a greater incidence of SVA than with the use of a double caval cannulation. It should be emphasized that although the present study supports the view that ischaemic injury of the atrium is the cause for postoperative arrhythmias, it does not reveal whether the source of the arrhythmias was the myocardium itself, dysfunction of the sinus node or both. Ischaemic injury of the atrium may not be the only cause of postoperative SVA. Cox [17] has suggested that a certain minimal rate of occurrence of atrial fibrillation is unavoidable since the incidence of atrial fibrillation after major surgical procedures not involving the heart is approximately 5% [18], a frequency similar to that shown in patients using intermittent ischaemia. If this is the toll that must be paid by the operation then one may be tempted to conclude that the arrhythmias seen in the group with intermittent ischaemia may not be attributed to the ischaemic insult to the atrium but to the surgery itself. Our results also suggest that although the type of cardioplegia does not influence the incidence of SVA during the postoperative period, it affects the persistence of atrial fibrillation during the follow up, with warm blood cardioplegia being less arrhythmogenic than cold crystalloid cardioplegia. These findings contrast with those reported by other investigators [16] showing that blood cardioplegia also reduces significantly the incidence of postoperative SVA when compared with crystalloid cardioplegia. Although subtle differences in the protocol of delivery of cardioplegia may explain the differences between these studies, the idea that ischaemic injury of the atrium is the primary cause of postoperative arrhythmias is strengthened. The preoperative withdrawal of b blockers has been shown to increase the incidence of postoperative
132
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
atrial arrhythmias [5]. This cannot account for the explanation of the differences observed in our study since all patients continued on their preoperative medications to the day of surgery. However, it cannot be ignored that b blockers and other pharmacological therapies (i.e., calcium antagonists, free radical scavengers, and so forth) may have an effect on the incidence of postoperative SVA via modulation of the tolerance of the atrium to ischaemia. Certain factors such as age [1], aortic cross-clamp time [2] and coronary artery disease [3] have been correlated with the development of postoperative SVA. However, our study shows that none of them alone (univariate analysis) or in combination (multivariate analysis) have a significant influence in the incidence of postoperative arrhythmias.
4.4. Limitation of the study and clinical implications A limitation of the present study is the diagnosis of arrhythmias by continuous ECG recording during the first 24 h after surgery and thereafter by clinical assessment at regular intervals and ECG recordings in suspected patients that may have underestimated their incidence. However, this limitation does not detract from the main findings of the study. This study provides evidence to suggest that the type of cardiac protection used during surgery determines the incidence of postoperative atrial arrhythmias. It also shows that postoperative atrial arrhythmia is not a benign condition and that they lead to a high incidence of residual atrial fibrillation during the follow up. Although our study does not show increased morbidity and mortality associated with the arrhythmia during the first 6 months after surgery, the persistence of atrial fibrillation may potentially put patients at higher risk of stroke and death in the long term. Future studies should establish the usefulness of antithrombotic therapies in patients presenting with residual atrial fibrillation following cardiac surgery.
References [1] Fuller JA, Adams GG, Buxton B. Atrial fibrillation after coronary artery bypass grafting: Is it a disorder of the elderly? J Thorac Cardiovasc Surgery 1989;97:821–5.
[2] Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993;56:539–49. [3] Kolvekar S, D’Souza A, Akhatar P, Reek C, Garratt C, Spyt T. Role of atrial ischaemia in development of atrial fibrillation following coronary artery bypass surgery. Eur J Cardio-Thoracic Surg 1997;11:70–5. [4] Mathew JP, Parks R, Savino JS et al. Atrial fibrillation following coronary artery bypass graft surgery predictors, outcomes, and resource utilization. J Am Med Assoc 1996;276:300–6. [5] Salazar C, Frishman W, Friedman S et al. Beta-blockade therapy for supraventricular tachyarrhythmias after coronary artery surgery: a propranolol withdrawal syndrome? Angiology 1979;30:816–9. [6] Pehkonen EJ, Reinikainen PM, Kataja MJ, Tarkka MR. Rhythm disturbances after blood and crystalloid cardioplegia in coronary artery bypass grafting. Scand J Thor Cardiovasc Surg 1995;29:23–8. [7] Solomon AJ, Kouretas PC, Hopkins RA, Katz NM, Wallace RB, Hannan RL. Early discharge of patients with new-onset atrial fibrillation after cardiovascular surgery. Am Heart J 1998;135:557– 63. [8] Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham study. Ann NY Acad Sci 1963;107:539–56. [9] Gajewski J, Singer RB. Mortality in an insured population with atrial fibrillation. J Am Med Assoc 1981;245:1540–4. [10] Krahn AD, Manfreda J, Tate RB, Mathewson FA, Cuddy TB. The natural history of atrial fibrillation: incidence, risk factors and prognosis in the Manitoba follow up study. Am J Med 1995;98:476– 84. [11] Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation 1998;98:946–52. [12] Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Arch Intern Med 1994;154:1449–57. [13] Yellon DM, Baxter GF, Garcia-Dorado D, Heusch G, Sumeray MS. Ischaemic preconditioning: present position and future directions. Cardiovasc Res 1998;37:21–33. [14] Mullen JC, Khan N, Weisel RD et al. Atrial activity during cardioplegia and postoperative arrhythmias. J Thorac Cardiovasc Surg 1987;94:558–65. [15] Butler J, Chong JL, Rocker GM, Pillai R, Westaby S. Atrial fibrillation after coronary artery bypass grafting: a comparison of cardioplegia versus intermittent aortic crossclamping. Eur J Cardiothorac Surg 1993;7:23–5. [16] Rousou JA, Meeran MK, Engelman RM, Breyer RH, Lemeshow S. Does the type of venous drainage or cardioplegia affect postoperative conduction and atrial arrhythmias. Circulation 1985;72(Suppl II):259–63. [17] Cox IL. A perspective of postoperative atrial fibrillation in cardiac operations. Ann Thorac Surg 1993;56:405–9. [18] Goldman L. Supraventricular tachyarrhythmias in hospitalized adults after surgery. Clinical correlates in patients over 40 years of age after major noncardiac surgery. Chest 1978;73:450–4.
Residual atrial fibrillation and clinical consequences following postoperative supraventricular arrhythmias Mahmoud Loubani FRCSI, Mark St. J. Hickey FRCSI, Tom J. Spyt FRCS, ˜ Manuel Galinanes FRCS* Division of Cardiac Surgery /Department of Surgery, University of Leicester, Glenfield Hospital, Groby Road, Leicester, LE3 9 QP, UK Received 18 October 1999; received in revised form 15 December 1999; accepted 28 January 2000
Abstract Aims: This retrospective study investigated whether the supraventricular arrhythmias (SVA) observed during cardiac surgery are limited to or persist beyond the postoperative period, their clinical consequences and whether they are influenced by preoperative and postoperative factors. Methods: A total of 375 patients undergoing elective bypass graft surgery over a 15-month period by three surgeons were included. All patients had their preoperative medications continued to the day of surgery and prophylactic anti-arrhythmic medications were not used in any of the cases. Standard anaesthetic techniques were used. Rhythm disturbances were diagnosed by ECG. The arrhythmias were treated medically or by cardioversion. All patients were followed up for 6 months. Results: Postoperative SVA occurred in 25% of patients. The commonest arrhythmia was atrial fibrillation (89.4%), followed by atrial flutter (6.4%) and supraventricular tachycardia (4.2%). In 89.8% of the cases, the arrhythmias occurred within the first four postoperative days with a maximum incidence on the second day (27.7%). Atrial fibrillation was still present in 50% of patients at hospital discharge and in 39% at 6 months follow up. Patients with arrhythmias had a prolonged hospital stay (7.762.6 vs. 6.062.6 days; P,0.05). There was no hospital mortality in the study and the incidence of postoperative stroke was equal in the sinus rhythm and arrhythmia patients (1.1%). SVA were more frequent when cardioplegia was used to protect the heart (32%) than with intermittent ischaemia (9%; P,0.001). At 6 months follow up, the patients receiving cardioplegia also had a higher prevalence of atrial fibrillation than those operated with intermittent ischaemia (41% vs. 22%; P,0.05). The incidence of SVA and persistence of atrial fibrillation was unrelated to other preoperative and intraoperative factors. Conclusion: Postoperative supraventricular arrhythmias have a long-lasting effect on cardiac rhythm: patients with SVA have a high probability of remaining in atrial fibrillation at hospital discharge and 6 months after surgery. The occurrence of atrial fibrillation seems to be influenced by the type of myocardial protection used but this does not appear to exert harmful effects. 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Atrial fibrillation; Cardiac surgery; Follow up
1. Introduction Supraventricular arrhythmias (SVA) are the most common complication following cardiac surgery; their incidence has been reported to vary from 10 to *Corresponding author. Tel.: 144-116-250-2449; fax: 144-116-2321282. ˜ E-mail address: [email protected] (M. Galinanes).
50% [1], and it is known that they are associated with increased morbidity and length of hospitalization [2], all adding to the cost of surgery. However, the prevalence and clinical consequences of these arrhythmias once the patient has been discharged from hospital are unknown. Furthermore, the cause of these rhythm disturbances still remains unclear, although a number of factors including age [1], anatomical distribution of coronary artery disease [3],
0167-5273 / 00 / $ – see front matter 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S0167-5273( 00 )00229-1
126
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
renal failure [1], poor left ventricular function [4], abrupt withdrawal of b blockers [5], history of rheumatic heart disease [2], chronic obstructive airway disease [2] and increasing aortic cross clamp time [2,6] have been reported to affect the incidence of SVA. The aims of the present study were to investigate whether postoperative SVA are short-lived and confined to the early postoperative period or persistent beyond hospital discharge, their clinical consequences, and whether their incidence is influenced by preoperative or intraoperative factors.
2. Materials and methods
2.1. Patients and exclusion criteria This was a retrospective study in which all patients (n5375) undergoing elective coronary artery bypass grafting surgery for a period of 15 months in a single institution and in normal sinus rhythm preoperatively were included. Data were collected when all the patients had a follow-up of at least 6 months. Patients operated on as emergency or undergoing a concomitant surgical procedure and those with preoperative history of SVA were excluded from the study. Patients were operated on by three surgeons each using only one of the three following methods of myocardial protection: cold crystalloid cardioplegia (n5122), normothermic blood cardioplegia (n5147) and intermittent ischaemia without cardioplegia (n5 106).
2.2. Medication and anaesthesia All preoperative medications, including b blockers and calcium antagonist were continued up to the day of surgery, except acetylsalicylic acid, which was discontinued at least one week before surgery. Patients received standard pre-medication and anaesthesia and no anti-arrhythmic medication (i.e., digoxin, amiodrone, b blockers) was administered prophylacticaly during or after surgery.
2.3. Surgery Cardiopulmonary bypass was established with a two-stage venous cannula in the right atrium and arterial return to the ascending aorta. In patients operated on with cold crystalloid cardioplegia, the body was cooled to 288C. The aorta was then crossclamped and cold (4–88C) crystalloid cardioplegia (St. Thomas’ cardioplegic solution no. 1) infused into the aortic root every 20 min. Once the distal anastomoses were completed, the aortic-cross clamp was removed, rewarming was initiated and the proximal anastomoses were completed. In patients with warm blood cardioplegia, body temperature was maintained at normothermia. The heart was then arrested with warm blood cardioplegia (KCl 20 mmol / l) infused into the aortic root and after completion of each distal anastomosis, additional doses of cardioplegia (KCl 10 mmol / l for the second to fourth dose and 7 mmol / l for all further doses) were given. The proximal anastomoses were carried out after releasing the aortic cross-clamp. In patients protected with intermittent ischaemia, the body was cooled to 328C and, in contrast with the other patients, no cardioplegia was used. Instead, the distal anastomoses were performed during brief periods of ischaemia achieved by cross-clamping the ascending aorta. The ischaemic episodes were interspaced by a period of reperfusion that ranged from 3 to 10 min, during which time the proximal anastomoses of the grafts to the ascending aorta were performed.
2.4. Detection of arrhythmia SVA was taken to include atrial fibrillation, atrial flutter and supraventricular tachycardia. Patients were monitored by continuous bedside electrocardiogram (ECG) monitoring of lead II for the first 24 h after surgery. The occurrence of SVA was confirmed by rhythm strip. Beyond this period the arrhythmias were detected by regular four-hourly clinical assessment or earlier if any symptoms were reported and confirmed by ECG. Peri-operative myocardial infarction was diagnosed with ECG changes and measurement of serum creatinine kinase (MB fraction). All patients were reviewed at 6 weeks and 6 months postoperatively in an outpatient clinic and questioned
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
with regard to residual angina. An ECG was done on each visit to evaluate the cardiac rhythm.
2.5. Treatment of arrhythmias Following the diagnosis of SVA, patients received one or a combination of the following therapies: digoxin, amiodarone or b blockers. If the medical therapy resulted in re-establishing sinus rhythm or in control of the ventricular rate and the patient was asymptomatic then the medication was continued for 6 weeks. At the end of this period and coinciding with the first postoperative visit at the outpatient clinic, if the rhythm had reverted to sinus then the medication was discontinued, if still in atrial fibrillation other medications were added or cardioversion was considered. Symptomatic patients or failure of anti-arrhythmic medications to control the ventricular rate led to cardioversion at any time during the follow-up. The procedure for the treatment of SVA followed by the three operating surgeons was identical. The persistence of the arrhythmia in spite of treatment was not a reason to maintain patients in hospital if ventricular heart rate was between 60 and 100 beats per min and this was not associated with hemodynamic impairment.
2.6. Expression of results and statistical analysis Data were expressed as means6standard deviation and the comparisons between the groups were made by the ANOVA analysis and x 2 test. Univariate and multivariate analyses were carried out to assess the significance of preoperative and intraoperative factors in the development of arrhythmias. A value of P, 0.05 was taken to be statistically significant.
3. Results
3.1. Postoperative supraventricular arrhythmias Table 1 shows that up to a quarter of patients undergoing elective bypass graft surgery have SVA in the postoperative period. It also shows that the clinical characteristics were comparable in patients with sinus rhythm and in those exhibiting SVA. Thus,
127
factors such as age, sex, angina class and the presence of hypertension, diabetes, hypercholesterolaemia or renal impairment have no significant effect. Furthermore, in the present study the distribution of coronary artery disease, the state of the left ventricular function, the ischaemic time (i.e., mean aortic cross-clamp time), incidence of perioperative myocardial infarction and persistence of angina at 6 months follow-up do not appear to influence the occurrence of postoperative SVA. However, Table 2 shows that the type of myocardial protection used during surgery was a significant factor in determining the occurrence of postoperative SVA. Patients receiving cold crystalloid cardioplegia and warm blood cardioplegia had a comparable incidence of SVA (.30%) that was higher than in patients operated on with the use of intermittent ischaemia (,10%). In all instances, atrial fibrillation was the commonest arrhythmia and accounted for almost 90% of SVA. Fig. 1 demonstrates that most of the arrhythmias occur within the first 6 days of the postoperative period with the majority of them (73.4%) being initiated within the first 3 days. Only two patients exhibited atrial fibrillation after this period; one required readmission for 3 days and the other was still an in-patient as a result of a wound infection. A total of seven patients were cardioverted in the immediate postoperative period and a further 17 in the ensuing 6 months. However, the choice of treatment of the rhythm disturbance whether it was one or more pharmacological agents or a combination with cardioversion did not seem to have any influence on the persistence of atrial fibrillation in our study patients.
3.2. Clinical consequences Patients with SVA had a significantly prolonged postoperative hospital stay (7.762.6 days) than patients who remained in sinus rhythm (6.062.6 days; P,0.001). This was due to the time required for the control of the ventricular rate with medical treatment or cardioversion. It should be noted, however, that maintenance of atrial fibrillation with controlled heart rate of less than 100 beats per min was not a reason for delaying hospital discharge. Interestingly, as seen in Table 3 only 50% of patients presenting SVA were
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
128
Table 1 Analysis of the clinical characteristics and operative data as risk factors for the occurrence of postoperative supraventricular arrhythmias a Supraventricular arrhythmia
Sinus rhythm
Total number of patients Age (years) Sex (male:female)
94 (25.5%) 63.568.2 76:15
281 (74.5%) 61.969.0 225:56
NS NS
NYHA angina class 1 and 2 3 and 4 Hypertension Diabetes Hypercholesterolaemia Renal impairment (Cr.120 mmol / l)
32 62 30 12 40 5
(34.0%) (64.9%) (33.0%) (13.2%) (42.5%) (5.3%)
78 (27.7%) 203 (72.4%) 106 (37.3%) 51 (18.0%) 128 (45.5%) 13 (4.6%)
NS NS NS NS NS NS
Coronary artery disease Single vessel disease Double vessel disease Triple vessel disease RCA disease LMS disease
3 23 68 63 17
(3.1%) (24.4%) (72.3%) (62.3%) (18.0%)
12 (4,2%) 64 (22.7%) 205 (72.9%) 246 (87.4%) 65 (23.1%)
NS NS NS NS NS
13 (4.6%) 41.6615.6
NS NS
12 (4.2%) 82 (29.1%) 125 (44.4%) 52 (18.5%) 9 (3.2%) 1 (0.4%)
NS NS NS NS NS NS
Poor LV (EF,30%) Mean aortic cross clamp time (mm)
3 (3.2%) 41.9614.2
Number of grafts 1 2 3 4 5 6
5 24 45 18 2 0
Peri-operative MI Residual angina at 6 months post-operative a
(5.3%) (25.5%) (47.8%) (19.1%) (19.1%)
1 (1.1%) 2 (2.1%)
4 (1.4%) 5 (1.7%)
P value
NS NS
RCA, right coronary artery; LMS, left main stem; LV, left ventricle; EF, ejection fraction; Cr, creatinine, MI, myocardial infarction.
in sinus rhythm at the time of hospital discharge. Moreover, in spite of antiarrhythmic treatment, 48% of the patients still remained in atrial fibrillation at 6 weeks follow up, a percentage that was slightly reduced to 39% after 6 months. Table 3 also shows that the percentage of patients in atrial fibrillation was
significantly higher in the cardioplegic groups than in the intermittent ischaemia group. Thus, atrial fibrillation was still present in between two-thirds and a half of patients protected with cardioplegia, whereas only one fifth of those operated on using intermittent ischaemia were in atrial fibrillation. It is worth noting
Table 2 Influence of the type of myocardial protection on postoperative supraventricular arrhythmias
Atrial fibrillation Atrial flutter Supraventricular tachycardia Overall incidence a b
Cold crystalloid cardioplegia
Warm blood cardioplegia
Intermittent ischaemia
Total
34 (27.8%) 3 (2.4%)
41 (27.9%) 3 (2.0%)
9 (8.5%)a 0b
84 (89.4%) 6 (6.4%)
3 (2.4%) 40 (32.8%)
1 (0.6%) 45 (30.6%)
0b 9 (8.5%)a
4 (4.2%) 94 (25.5%)
P,0.001 vs. both cardioplegia groups. P,0.05 vs. both cardioplegia groups.
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
129
Fig. 1. Day of onset of supraventricular arrhythmia (SVA) following cardiac surgery.
that although the incidence of atrial fibrillation was similar in patients protected with cold crystalloid cardioplegia and those receiving warm blood cardioplegia at the time of hospital discharge, a significantly lower percentage of individuals remained in atrial
fibrillation in the latter group than in the former at 6 weeks and 6 months of follow-up. In considering the high prevalence of atrial fibrillation even after 6 months since operation, it is worth noting that with the exception of one patient developing a stroke in
130
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
Table 3 Persistence of atrial fibrillation a Atrial fibrillation
Cold crystalloid cardioplegia
Warm blood cardioplegia
Intermittent ischaemia
At hospital discharge At 6 weeks follow up At 6 months follow up Overall incidence
21 25 20 40
25 18 15 45
1 2 2 9
(52.5%) (62.5%) (50.0%) (32.8%)
(55.5%) (40.0%)c (33.3%)c (30.6%)
(1.1%)b (22.2)b (22.2%)b (8.5%)b
Total 47 (50%) 45 (47.8%) 4 (4.2%) 94 (25.5%)
a
The figures in brackets express the percentage of patients with atrial fibrillation of those that exhibited postoperative supraventricular arrhythmias. P,0.001 vs. both cardioplegia groups. c P,0.05 vs. cold crystalloid cardioplegia. b
the third postoperative day there were no further complications associated with arrhythmias. The incidence of stroke in patients with sinus rhythm was equal to that of patients with SVA (1.1%).
4. Discussion The present study has shown for the first time that more than a third of patients exhibiting SVA following bypass graft surgery still remain in atrial fibrillation after 6 months follow up and that this occurs in spite of the continuation of anti-arrhythmic medication. In addition, our study provides convincing evidence that the type of myocardial protection determines not only the incidence of postoperative SVA but also the prevalence of atrial fibrillation within the first 6 months after surgery. These results raise a number of important points that warrant further discussion.
4.1. Postoperative supraventricular arrhythmias SVA is the most common complication following bypass graft surgery and this is supported by our study. This shows that the overall incidence of SVA is 25% of elective graft surgery patients that did not have a previous history of rhythm disturbances, which compares with other studies reporting an incidence oscillating between 10 and 50% [1]. In this selected group of patients, SVA lengthened hospitalization but it did not increase morbidity. This contrasts with other studies including patients with graft and valve surgery alone and in combination, and patients with heart transplantation [2] in which SVA were associated to an increased incidence of stroke, increased occurrence of ventricular arrhythmias and a
greater need for implantation of pacemakers. In a recent study, Solomon et al. [7] have reported that no thromboembolic events occurred after cardiac surgery in 12 patients that were discharged in atrial fibrillation and were followed up for over 6 months. However, in their study all patients received warfarin anti-coagulation. In the present study, only one patient with SVA had a stroke in the third postoperative day and this was possibly attributed to the cardiac arrhythmia since other potential cardiac and vascular causes were ruled out. The distinction of the study population is important because SVA in patients with other types of surgery involving the implantation of intravascular foreign materials and with increased cardiac cavities dimensions and altered flow patterns and hemodynamics may be more predisposed to cardiovascular complications than patients undergoing graft surgery alone.
4.2. Residual atrial fibrillation The most striking finding of this study is that half of the patients experiencing postoperative SVA leave the hospital in atrial fibrillation and that a high proportion of these remain with the arrhythmia after 6 weeks and 6 months follow up. Furthermore, patients that do not develop SVA during the early postoperative period are unlikely to manifest with atrial fibrillation during the ensuing follow up. The high prevalence of atrial fibrillation after hospital discharge in patients with postoperative SVA may have important clinical implications since individuals in atrial fibrillation are at high risk of stroke [8] and death [9–11]. The absence of morbidity and mortality attributable to atrial fibrillation after hospital discharge in our study may be related to the insufficient number of patients entered in the study
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
and to the short follow up period. Another possibility is that all patients received warfarin and / or aspirin (150 mg / day) for the entire study period and this per se may have reduced the risk of stroke. The thesis that anti-platelet or anti-coagulation therapy may explain the results of the present study are supported by placebo-controlled trials in which warfarin was shown to substantially reduce stroke rates [12]. Clearly, further studies are required to define the clinical repercussions of the persistence of atrial fibrillation after cardiac surgery and to determine the best therapeutic strategies for its management.
4.3. Possible causes Our study demonstrates that both the incidence of postoperative SVA and the residual atrial fibrillation beyond the postoperative period depend on the type of myocardial protection used. Thus, the occurrence of arrhythmia was three times higher in patients protected with both blood and crystalloid cardioplegia than in those operated with short intermittent periods of ischaemia, a method that mimics the potent protective protocols of ischaemic preconditioning [13]. These results provide support to the thesis that ischaemic injury during cardiac surgery may be responsible for the development of postoperative arrhythmia and at first sight, one may be tempted to conclude that intermittent ischaemia is more effective than either blood or crystalloid cardioplegia. However, it may be possible that atrial and ventricular myocardium are differently protected during cardiac surgery. In this connection, it has been reported that atrial myocardium is less effectively cooled than ventricular myocardium when hearts are infused with hypothermic cardioplegia [14]. If this is the case, then it may be expected to have a greater degree of ischaemic injury in the atrium than in the ventricle and therefore an increased susceptibility to the development of postoperative SVA when cardioplegia is the protective modality used. In a previous study, Butler et al. [15] showed that intermittent ischaemia reduced the incidence of postoperative SVA from 25.6% in patients protected with hypothermic crystalloid cardioplegia to 19.3%; however the difference did not achieve statistical significance. In this study [15], a two-stage venous cannula was used to establish cardiopulmonary bypass in the cardioplegia
131
group whereas a basket cannula located in the right atrium was utilised in the intermittent ischaemia group, and this may have accounted for the higher incidence of SVA in patients treated with intermittent ischaemia in their study. Certainly, the use of a basket cannula will allow a major contact of blood with the atrium and probably a greater distension of the cavity that may increase the ischaemic injury of the atrium. This possibility is supported by the findings of Rousou et al. [16] that the use of a single cannula is associated with less effective cooling of the right atrium and a greater incidence of SVA than with the use of a double caval cannulation. It should be emphasized that although the present study supports the view that ischaemic injury of the atrium is the cause for postoperative arrhythmias, it does not reveal whether the source of the arrhythmias was the myocardium itself, dysfunction of the sinus node or both. Ischaemic injury of the atrium may not be the only cause of postoperative SVA. Cox [17] has suggested that a certain minimal rate of occurrence of atrial fibrillation is unavoidable since the incidence of atrial fibrillation after major surgical procedures not involving the heart is approximately 5% [18], a frequency similar to that shown in patients using intermittent ischaemia. If this is the toll that must be paid by the operation then one may be tempted to conclude that the arrhythmias seen in the group with intermittent ischaemia may not be attributed to the ischaemic insult to the atrium but to the surgery itself. Our results also suggest that although the type of cardioplegia does not influence the incidence of SVA during the postoperative period, it affects the persistence of atrial fibrillation during the follow up, with warm blood cardioplegia being less arrhythmogenic than cold crystalloid cardioplegia. These findings contrast with those reported by other investigators [16] showing that blood cardioplegia also reduces significantly the incidence of postoperative SVA when compared with crystalloid cardioplegia. Although subtle differences in the protocol of delivery of cardioplegia may explain the differences between these studies, the idea that ischaemic injury of the atrium is the primary cause of postoperative arrhythmias is strengthened. The preoperative withdrawal of b blockers has been shown to increase the incidence of postoperative
132
M. Loubani et al. / International Journal of Cardiology 74 (2000) 125 – 132
atrial arrhythmias [5]. This cannot account for the explanation of the differences observed in our study since all patients continued on their preoperative medications to the day of surgery. However, it cannot be ignored that b blockers and other pharmacological therapies (i.e., calcium antagonists, free radical scavengers, and so forth) may have an effect on the incidence of postoperative SVA via modulation of the tolerance of the atrium to ischaemia. Certain factors such as age [1], aortic cross-clamp time [2] and coronary artery disease [3] have been correlated with the development of postoperative SVA. However, our study shows that none of them alone (univariate analysis) or in combination (multivariate analysis) have a significant influence in the incidence of postoperative arrhythmias.
4.4. Limitation of the study and clinical implications A limitation of the present study is the diagnosis of arrhythmias by continuous ECG recording during the first 24 h after surgery and thereafter by clinical assessment at regular intervals and ECG recordings in suspected patients that may have underestimated their incidence. However, this limitation does not detract from the main findings of the study. This study provides evidence to suggest that the type of cardiac protection used during surgery determines the incidence of postoperative atrial arrhythmias. It also shows that postoperative atrial arrhythmia is not a benign condition and that they lead to a high incidence of residual atrial fibrillation during the follow up. Although our study does not show increased morbidity and mortality associated with the arrhythmia during the first 6 months after surgery, the persistence of atrial fibrillation may potentially put patients at higher risk of stroke and death in the long term. Future studies should establish the usefulness of antithrombotic therapies in patients presenting with residual atrial fibrillation following cardiac surgery.
References [1] Fuller JA, Adams GG, Buxton B. Atrial fibrillation after coronary artery bypass grafting: Is it a disorder of the elderly? J Thorac Cardiovasc Surgery 1989;97:821–5.
[2] Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993;56:539–49. [3] Kolvekar S, D’Souza A, Akhatar P, Reek C, Garratt C, Spyt T. Role of atrial ischaemia in development of atrial fibrillation following coronary artery bypass surgery. Eur J Cardio-Thoracic Surg 1997;11:70–5. [4] Mathew JP, Parks R, Savino JS et al. Atrial fibrillation following coronary artery bypass graft surgery predictors, outcomes, and resource utilization. J Am Med Assoc 1996;276:300–6. [5] Salazar C, Frishman W, Friedman S et al. Beta-blockade therapy for supraventricular tachyarrhythmias after coronary artery surgery: a propranolol withdrawal syndrome? Angiology 1979;30:816–9. [6] Pehkonen EJ, Reinikainen PM, Kataja MJ, Tarkka MR. Rhythm disturbances after blood and crystalloid cardioplegia in coronary artery bypass grafting. Scand J Thor Cardiovasc Surg 1995;29:23–8. [7] Solomon AJ, Kouretas PC, Hopkins RA, Katz NM, Wallace RB, Hannan RL. Early discharge of patients with new-onset atrial fibrillation after cardiovascular surgery. Am Heart J 1998;135:557– 63. [8] Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham study. Ann NY Acad Sci 1963;107:539–56. [9] Gajewski J, Singer RB. Mortality in an insured population with atrial fibrillation. J Am Med Assoc 1981;245:1540–4. [10] Krahn AD, Manfreda J, Tate RB, Mathewson FA, Cuddy TB. The natural history of atrial fibrillation: incidence, risk factors and prognosis in the Manitoba follow up study. Am J Med 1995;98:476– 84. [11] Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation 1998;98:946–52. [12] Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Arch Intern Med 1994;154:1449–57. [13] Yellon DM, Baxter GF, Garcia-Dorado D, Heusch G, Sumeray MS. Ischaemic preconditioning: present position and future directions. Cardiovasc Res 1998;37:21–33. [14] Mullen JC, Khan N, Weisel RD et al. Atrial activity during cardioplegia and postoperative arrhythmias. J Thorac Cardiovasc Surg 1987;94:558–65. [15] Butler J, Chong JL, Rocker GM, Pillai R, Westaby S. Atrial fibrillation after coronary artery bypass grafting: a comparison of cardioplegia versus intermittent aortic crossclamping. Eur J Cardiothorac Surg 1993;7:23–5. [16] Rousou JA, Meeran MK, Engelman RM, Breyer RH, Lemeshow S. Does the type of venous drainage or cardioplegia affect postoperative conduction and atrial arrhythmias. Circulation 1985;72(Suppl II):259–63. [17] Cox IL. A perspective of postoperative atrial fibrillation in cardiac operations. Ann Thorac Surg 1993;56:405–9. [18] Goldman L. Supraventricular tachyarrhythmias in hospitalized adults after surgery. Clinical correlates in patients over 40 years of age after major noncardiac surgery. Chest 1978;73:450–4.