Background There has been an increasing number of children with congenital heart disease that undergo primary or second systemic-pulmonary shunt, while there are few reports on the second systemic-pulmonary shunt. Therefore, this study summarizes the experience of second systemic-pulmonary shunt for congenital heart disease in our hospital. Methods and results Sixty-five children with congenital heart disease who underwent systemic-pulmonary shunt for the second time in our hospital were analyzed. At the early stage after the operation, cyanosis improved and SpO2 significantly increased. One patient died in hospital (1.54%) and the causes of death were aggravated atrioventricular regurgitation, low cardiac output syndrome, and liver failure. Early complications occurred in 18 patients (27.7%). All the children were rechecked in our hospital every 3–6 months and the McGoon index significantly increased. Conclusion Systemic-pulmonary artery shunt can promote pulmonary vascular development, improve cyanosis symptoms, and increase the chance of radical treatment in children with pulmonary vascular dysplasia.

10mm (1%, n=1). Locations of para-device leaks identified were inferior (11%, n=8), superior (15%, n=11), non-specific (7%, n=5). In addition, one patient had device thrombus. Incidence of PFO pre-Watchman was 16%. There was a 17% (n=8) incidence of de novo IAS defects post-Watchman implant after comparison with pre-procedural TOE. Mean IAS defect size pre-and post-Watchman was 0.12±0.38 vs 0.29±0.58 cm, respectively. This change in IAS defect size of 0.13±0.57 trended towards significance (p=0.09). Conclusions: A modest incidence of para-device leaks and rare but significant incidence of para-device thrombus was observed by TOE post-Watchman implantation. Further work is required to correlate to potential implications of these findings on longer term outcomes related to recurrent strokes. The significance of persistent IAS defects post device implant also requires further investigation in longitudinal registry follow-up. P6135 | BEDSIDEThe role of left atrial strain and procollajen type III n-terminal peptide levels in determination of paroxysmal atrial fibrillation in ischemic stroke S. Eroglu, K. Akyol, E. Ciftci, E. Sade, H. Muderrisoglu. Baskent University, Faculty of Medicine, Ankara, Turkey Introduction and aim: Paroxysmal atrial fibrillation (PAF) is an important etiological cause of ischemic stroke. Although rhythm recorders are often used there are still difficulties in detecting PAF. Procollagen type III N-terminal peptide (P3NP) is associated with cardiac fibrosis. So we aimed to investigate the role of left atrial strain and P3NP levels in the detection of PAF in patient with ischemic stroke. Methods: All of the 100 patients (55 men, mean age 68.5±14.3 years) with acute ischemic stroke were included. Patients with acute ischemic stroke whose basal rhythm was sinus but had a spontaneous termination of atrial fibrillation detected by 24 hour holter monitorization were considered as PAF groups. Patients with acute ischemic stroke who did not have detected PAF by holter were accepted as the control group. Left atrial strain analysis was performed using an analysis package (EchoPAC, General Electric, Horten,Norway) based on a speckle-tracking principle. Peak atrial systolic strain, strain rate, peak early and late diastolic strain rate were measured. Serum P3NP levels were measured by ELISA methods. Results: Left atrial volume was increased in PAF group than in controls (54,6±13,4mL vs 46,2±1 0,9mL P=0,001). Left atrial peak systolic strain was significantly lower in patient with PAF than those with controls (P<0,001 Table 1). Similarly, left atrial peak early diastolic strain rate measurements, were also found to be lower in the PAF group than in the controls (p<0.05 Table 1). P3NP levels in patients with PAF were significantly higher than those in the control group (27768,5±47703.5 pg/ml vs 13328,4±21299.9 pg/ml, p<0,05). Left atrial peak systolic strain was 26.5% with 82% sensitivity and 62% specificity [receiver operating characteristic area 0,778, 95% confidence interval (0,686-0,871)] as indicatin...