Introduction Data from cardiac CT (CCT) is often fused with eletrophysiological maps to allow precise pulmonary vein isolation and can detect left atrial appendage (LAA) thrombus in patients undergoing radio-frequency AF ablation (RFA). We present a complete audit cycle that reduced the prevalence of pseudo-thrombus (caused by poor initial mixing of contrast) in the LAA on CCT following implementation of a new protocol.

Methods A standard of 100% positive and negative predictive value for the detection of LAA thrombus in patients undergoing CT assessment pre AF ablation was set. 225 consecutive patients were included with 7 patients excluded due to congential heart disease or previous cardiac surgery. 90 consecutive scans included in the initial audit. The diagnostic accuracy (sensitivity, specificity, positive (PPV) and negative predictive value (NPV)) for LAA thrombus detection was determined. A new 2-phase acquisition protocol (additional 60 second delayed scan) was introduced and the subsequent 128 consecutive cases were re-audited to determine if the standard had been met and compared to the published range in the literature. All patients underwent a TOE (as part of the standard clinical work-up) as the gold-standard reference. The additional radiation burden following the introduction of the delayed scan was determined.

Results A total of 8 true LAA thrombi were detected (5 and 3 using standard acquisition and 2-phase, respectively). The use of the 2-phase acquisition significantly reduced pseudo-thrombus detection (30 vs. 0 cases; p < 0.0001) improving the specificity and PPV (Table 1). The mean additional radiation burden was 0.4 (0.2) mSv.

Conclusion Implementation of an additional 60 second delayed scan improved the detection of LAA thrombus in patients referred for RFA potentially negating the need for a separate TOE for this purpose. This incurs a minimal additional radiation dose and requires no additional contrast.