as non-CTEPH and three MRI were considered non-diagnostic. MR lung perfusion was not performed in 11 patients due to impaired renal function or patient unable to tolerate breathhold. None of the surgically accessible disease was missed with either modality. The MR lung perfusion had a sensitivity of 96%, specificity of 92%, PPV of 87%, NPV of 98% and accuracy of 94% for diagnosing CTEPH. Perfusion scintigraphy had a sensitivity of 93%, specificity of 90%, PPV of 83%, NPV of 96% and accuracy of 91%. Conclusion Our results show that MR lung perfusion compared to 4-view static perfusion scintigraphy has an overall higher accuracy compared to perfusion scintigraphy and has a role in assessment of patients with suspected CTEPH. Background The incidence of pulmonary hypertension (PH) after an episode of acute pulmonary embolism (PE) is thought to be up to 4%. The incidence of persistent clot without PH is less clear. We evaluated the incidence of persistent perfusion defects in patients followed up after PE and assessed the impact on pulmonary arterial pressure measured by echocardiography. Methods The clinical pathway for outpatient follow-up of patients with PE includes a nuclear medicine ventilationperfusion scan at 5 months post PE. When this is positive, an echocardiogram is requested to look for evidence of PH. A retrospective study of consecutive patients attending outpatient follow-up was carried out to determine the incidence of persistent perfusion defects and echocardiographic findings suggestive of PH. Results Ninety-three patients were identified between February 2009 and July 2011 and their clinical data were studied retrospectively. 59 patients (63%) had persistent perfusion defects at 6 months. Of these 49 (83%) underwent echocardiography of which 12 had undetectable tricuspid regurgitation (TR). In the remaining 37, pulmonary artery systolic pressure (PASP) could be estimated from the TR velocity. Results are shown in Abstract S24 figure 1 and are skewed due to the presence of patients with PH. Median PASP (with IQR) was 32 mm Hg (26e39). The shaded area shows the 95% CIs for the normal PASP (2869.8 mm Hg, mean 6 2SD). Fourteen patients had estimated PASP above the normal range. Six patients were investigated with cardiac catheterisation S24

No abstract