RÉSUMÉ
<p><b>INTRODUCTION</b>The device closure of atrial septal defects has evolved over the years. In the early days of transcatheter occlusion, balloon sizing was used to choose an appropriate sized device. We postulate that balloon sizing does not value-add to the procedure and is unnecessary.</p><p><b>MATERIALS AND METHODS</b>Patients who had balloon sizing, with (Group 1, n = 38) or without (Group 2, n = 21) atrial septal defect closure, were compared to another group (Group 3, n = 64) who had atrial septal defect closure without balloon sizing. Although the atrial septal defect size (mm) in those without balloon sizing (Group 3) compared to patients who had balloon sizing (Group 1) (18.3 +/- 5.4 vs 14.8 +/- 5.8; P = 0.021) was larger, the Amplatzer septal occluder size chosen (mm) (21.6 +/- 6.3 vs 21.2 +/- 8.1; P = 0.693) was similar.</p><p><b>RESULTS</b>We analysed the degree of absolute sizing, defined as [(Balloon or Amplatzer occluder size) - (transoesophageal echocardiography size)], versus relative sizing, which is defined as [(Balloon or Amplatzer occluder size)--(transoesophageal echocardiography size) / (Balloon or Amplatzer occluder size)]. It was evident that there was greater absolute and relative over-sizing (6.3 +/- 4.4 mm vs 4.2 +/- 2.1 mm; P = 0.009 and 28.3 +/- 15.4% vs 20.0 +/- 7.0%; P = 0.001, respectively) in patients with balloon sizing (Group 1) compared to those who did not (Group 3). Even a greater degree of absolute (5.1 +/- 3.9 mm vs 9.5 +/- 4.7 mm; P <0.001) and relative over-sizing (24.8 +/- 15.6% vs 33.0 +/- 13.6%; P = 0.001) was observed in patients who had balloon sizing but there was no closure (Group 2) compared to those who had balloon sizing and closure of their defects (Group 1).</p><p><b>CONCLUSION</b>Our results showed that balloon sizing tended to over-size the atrial septal defect. This may have an important bearing in selecting a larger device than necessary, or even precluding transcatheter closure of the larger atrial septal defects. It is also associated with increased procedural, fluoroscopy time and cost. We suggest that balloon sizing may no longer be necessary in the protocol of device closure of an atrial septal defect.</p>