Comparison of Doppler echocardiography and cardiac catheterization in patients requiring valve surgery: search for a 'gold standard'

Abstract

Objective: To compare the sensitivities of Doppler echocardiography and cardiac catheterization in the diagnosis of severe valvular heart disease in patients requiring valve surgery.

Design: Retrospective analysis of Doppler echocardiograms and cardiac catheterizations.

Setting: Tertiary referral cardiovascular centre in a university setting.

Patients: Sixty-nine patients undergoing valve surgery between July 1988 and July 1990.

Results: The sensitivities of echocardiography and cardiac catheterization were 84 and 87%, respectively (P = 1.0) in 32 patients who underwent aortic valve surgery primarily for severe aortic stenosis; 83 and 67%, respectively (P = 1.0) in six patients with severe aortic regurgitation, and 100 and 85%, respectively (P = 1.0) in seven patients with combined severe aortic stenosis and regurgitation. The sensitivities of echocardiography and cardiac catheterization in 11 patients who underwent mitral valve surgery for severe mitral stenosis were 73 and 91%, respectively (P = 0.6) and 69 and 92%, respectively (P = 0.3) in 13 patients with severe mitral regurgitation. Sensitivities of echocardiography and cardiac catheterization in the diagnosis of severe tricuspid regurgitation in five patients who had tricuspid valve repair were 100 and 80%, respectively (P = 1.0). Two patients with severe aortic stenosis by echocardiography, but not by catheterization, did not undergo aortic valve replacement during valvular surgery; both required aortic valve replacement within two years of initial surgery because of heart failure. Four patients with severe tricuspid regurgitation identified by echocardiography did not have tricuspid repair; three had pulmonary hypertension and these patients had resolution of tricuspid regurgitation on follow-up. One patient with severe tricuspid regurgitation and absence of pulmonary hypertension required reoperation for tricuspid valve repair 10 months after initial operation.

Conclusions: The sensitivity of echocardiography and cardiac catheterization in the detection of severe valvular lesions requiring surgery is similar. Discordant results should be reviewed carefully with knowledge of the inherent pitfalls of both techniques in order to ensure optimal patient outcome.