Diagnostic accuracy of multidetector computed tomography in the detection of left atrial/left atrial appendage thrombus: a meta-analysis

Background and objective The current results on the diagnostic accuracy of multidetector computed tomography (MDCT) in the detection of left atrial/left atrial appendage (LA/LAA) thrombus are conflicting. Aim The aim of the present study was to determine the diagnostic accuracy of MDCT in LA/LAA thrombus with meta-analysis. Methods We searched for studies in PubMed, Embase and Cochrane library prior to May 2012 evaluating the accuracy of MDCT in detecting LA/LAA thrombus. Primary results were summarised using a random-effects model or a fixed-effects model. Receiver operating characteristic curves were used to summarise overall diagnosis accuracy. Metaregression and subgroup analysis were used to explore the potential sources of heterogeneity. Results A total of 10 studies with 1313 subjects was included in this meta-analysis. The summary estimates of sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, diagnostic odds ratio and the area under cure of overall analysis were 0.84, 0.93, 9.32, 0.21, 50.84 and 0.951, respectively, but all with significant heterogeneity (P < 0.01). Meta-regression and subgroup analysis showed that using electrocardiogram (ECG)-gating technique was a source of heterogeneity (P = 0.022); studies using ECG-gating technique had a higher summary sensitivity than studies with non-gating technique (0.97 vs 0.33). Conclusion Our results suggest that MDCT is a potentially useful technique in the diagnosis of LA/LAA thrombus, especially when ECG gating is applied.