The hypothesis being studied is that add-on arrhythmia surgery in patients with atrial fibrillation (AF) undergoing valvular or coronary surgery improves quality of life, is cost-effective, reduces perioperative and long-term morbidity associated with AF.

Atrial fibrillation (AF) is connected with an increased morbidity and mortality. In addition, quality of life is diminished due to palpitations, dyspnea, dizziness and syncope. AF is frequently associated with valvular and coronary disease. In the AF patients undergoing valvular or coronary surgery the arrhythmia almost always relapses. For symptom control anti-arrhythmic drugs and cardioversion are used but breakthrough arrhythmias and side effects of the drugs happen frequently. For more effective symptom control "add-on" arrhythmia surgery is being advocated. However, at present we do not know whether add-on arrhythmia surgery indeed affects morbidity and quality of life. In this respect the benefit of chronic sinus rhythm has to outweigh the risks of a prolonged operation. In addition, cardiovascular complaints unrelated to AF may persist even after successful operation, thus offsetting the benefit of chronic sinus rhythm. Add-on surgery is more costly than standard surgery but this may compare favourably with shorter hospital admission due to less frequent post-operative AF.

Valvular heart disease is frequently associated with ventricular remodelling: a decreased ventricular function and atrial dilatation. AF itself may worsen heart failure due to a tachycardiomyopathy. Elimination of AF might therefore enhance recovery from structural and functional remodelling and promote recovery of quality of life after the operation.

The PIAF, RACE and AFFIRM have shown that chronic sinus rhythm is not necessarily associated with a reduced morbidity or enhanced quality of life. PIAF however showed that exercise tolerance was better when rhythm control was achieved. Further analyses of RACE and AFFIRM are pending. One drawback of the above studies is the fact that chronic sinus rhythm is difficult to obtain. In PIAF, RACE and AFFIRM only 30 to 50% was in sinus rhythm at the end of follow-up. By contrast, arrhythmia surgery is a highly effective treatment in this respect.

Forty patients underwent a (phase 1-study) coronary bypass- or valve surgery with add-on arrhythmia surgery in the same way as in this protocol proposed. This means epicardial on beating heart and without use of the heart-lung machine. In the last follow-up 80 % of the patients not longer were in atrial fibrillation. With similar treatment procedures, but more invasive, so on the arrested heart and endo cardially, success percentages reported varying from 60 % till 80%.

This large variance in success rate is probably related to the primary course of the disease and the degree of the morphological abnormality. In spite of these meaningful results ''add-on'' arrhythmia surgery is no general accepted treatment. The intended patient population remains generally untreated. Historical data of patients from the university hospital of Maastricht show that no add on treatment has a success rate of 25% of patients in sinus rhythm.

Considering the above a randomised comparison of add-on arrhythmia surgery and standard surgery is warranted.

Inclusion Criteria:

1. Patients undergo valvular surgery, coronary surgery, or a combination of both.
2. All patients have documented chronic atrial fibrillation of paroxysmal atrial fibrillation.
3. Patients have given written informed consent.

Exclusion Criteria:

1. Patients who do not speak Dutch or can not read Dutch.
2. Patients with a Sick Sinus Syndrome.
3. Patients with contraindications for oral anticoagulant agents.
4. Patients unable to express their specific wishes.