Atrial fibrillation: rhythm control does not clearly improve symptoms and increases hospital admissions and adverse effects compared with rate control

Clinical bottom line (level 1b)

  1. Patients with persistent atrial fibrillation who received rhythm control with amiodarone and DC cardioversion compared with rate control with diltiazem had better exercise tolerance, but did not clearly have an improvement in symptoms or quality of life.
  2. Patients on rhythm control were more likely to require hospitalisation (NNH = 2 at 12 months) , more likely to develop medication-related adverse effects (NNH = 6 at 12 months) and have to discontinued medication (NNH = 9 at 12 months) .
Hohnloser et al: Lancet 2000; 356 : 1789-1794
Expires November 2003

The study

Unblinded ?concealed randomised trial with intention-to-treat
Setting: 21 centres, Germany

252 patients (aged mean 61, 73% male) persistent and symptomatic atrial fibrillation (for 7 to 360 days)

Excluded if
  • pregnancy
  • pacemaker therapy
  • contraindication to systemic anticoagulation therapy
  • intracardiac thrombus on echocardiogram, or central or peripheral embolisation within previous 3 months
  • hypertrophic cardiomyopathy
  • amiodarone therapy within previous 6 months
  • acute thyroid dysfunction
  • AF with rate < 50 beats/min
  • known sick sinus syndrome
  • AF in the setting of Wolff-Parkinson-White syndrome
  • CABG or valve replacement within previous 3 months
  • aged < 18, > 75
  • congestive heart failure, NYHA class IV
  • unstable angina, or myocardial infarction within previous 30 days

    Control Group: (n = 125, 125 analysed): rate control using diltiazem 90 mg two to three times a day, plus additional therapy if required
    Experimental Group: (n = 127, 127 analysed): rhythm control: pharmacological cardioversion using amiodarone 600 mg for 3 weeks, followed by DC cardioversion if required. Sinus rhythm was maintained using amiodarone 200 mg daily.
    All patients were anticoagulated with a target INR 2.0 to 3.0
    98% followed for 12 months

    The evidence

    Outcome Time to outcome CER EER RRR
    (95% CI)    		 ARR
    (95% CI)    		 NNT
    (95% CI)
    atrial fibrillation 12 months 112
    (89.6%)    		 56
    (44.1%)    		 51%
    (40% to 60%)    		 45.5%
    (35.4% to 55.7%)    		 2
    (2 to 3)
    hospital admission 12 months 30
    (24.0%)    		 87
    (68.5%)    		 -190%
    (-300% to -100%)    		 -44.5%
    (-55.5% to -33.5%)    		 -2
    (-3 to -2)
    at least one adverse effect 12 months 58
    (46.4%)    		 80
    (63.0%)    		 -36%
    (-71% to -8%)    		 -16.6%
    (-28.7% to -4.47%)    		 -6
    (-22 to -3)
    discontinuation of medication due to adverse effects 12 months 17
    (13.6%)    		 31
    (24.4%)    		 -79%
    (-210% to -5%)    		 -10.8%
    (-20.4% to -1.22%)    		 -9
    (-82 to -5)

  • There was no significant difference in symptoms (palpitations, dyspnoea, dizziness) nor quality of life reported between the two groups at any time.
  • Patients who had rhythm control performed better on a 6 minute walking test than those who have rate control after 12 weeks (p = 0.012)

    Comments

    1. Patients were stratified by centre, and randomised in blocks of six.
    2. Two patients in each group died during follow-up.
    3. Patients in the rhythm control group were mainly admitted to hospital for DC cardioversion (67%) or due to amiodarone-related adverse effects (27%).

    Citation

    1. Hohnloser SH, Kuck KH, Lillienthal J, et al: rhythm or rate control in atrial fibrillation - pharmacological intervention in atrial fibrillation (PIAF): a randomised trial. Lancet 2000; 356 : 1789-1794
    Search Terms: from EBM journal
    Contributor: Chris Ball, November 2001
    Reviewer:

    Clinical Question.
    Patient persistent atrial fibrillation
    Intervention or Exposure rhythm control
    Comparison rate control
    Outcome symptoms, adverse effects, hospital admission