Is cardiac magnetic resonance imaging as accurate as echocardiography in the assessment of aortic valve stenosis?

A best evidence topic was written according to a structured protocol. The question addressed was: is cardiac magnetic resonance (CMR) imaging as accurate as echocardiography in the assessment of aortic valve stenosis? Altogether 239 papers were found using the reported search. Only 12 demonstrated the best evidence to answer the clinical question. Nine of these 12 papers found CMR to correlate well with transthoracic echocardiography (TTE) or transoesophageal echocardiography (TOE) in the evaluation of aortic valve stenosis. When aortic valve areas were measured with cardiac tomography (CT) or cardiac catheterization (CC), four papers found CMR to be more accurate than TTE. Eight of 12 papers found CMR to have excellent reliability and reproducibility, as demonstrated by the low inter- and intraobserver variability. Four papers did not estimate intra- or interobserver variability. One paper noted a sensitivity and specificity of 96 and 100%, respectively, when using CMR to detect severe aortic stenosis (AS) that had been diagnosed during CC. A second paper noted a lower sensitivity and specificity of 78 and 89%, respectively, but this was still better than the sensitivities and specificities found when using TOE or TTE to detect severe AS, as noted on CC. We conclude that current evidence finds echocardiography and CMR to be equally reliable in assessing aortic stenosis. CMR has better inter- and intraobserver reliability and demonstrates an advantage over echocardiography in the detection of severe AS with greater specificity and sensitivity. The final choice, however, is as likely to be influenced by the availability of magnetic resonance imaging and expertise in interpreting the results as by accuracy and reliability.