Estimation of mitral valve area from regression analysis of the pressure gradient in mitral stenosis

Estimation of mitral valve area (MVA) in the cardiac catheterization laboratory is prone to pitfalls because of the time required for calculations and inaccuracies in the measurement of cardiac output. Because the rate of decrease in the mitral gradient directly correlates with the severity of mitral stenosis, an on-line estimate of MVA at the time of catheterization may be possible with regression analysis of digitized pressure recordings. A total of 61 comparisons of mitral gradient measurements and MVA were obtained in 37 patients at diagnostic catheterization and in 24 patients after balloon mitral valvotomy. Linear and nonlinear regression parameters yielded pressure half-time values and empiric constants similar to those used in Doppler echocardiography for estimation of MVA. The correlations derived from linear analysis were as good as those obtained from nonlinear analysis: from linear analysis, MVAregression = 0.79.MVAGorlin -0.03; r2 = 0.64, p = 0.0001; and from double exponential analysis, MVAregression = 0.86.MVAGorlin -0.07; r2 = 0.74; p = 0.0001. The correlations were not significantly affected by the presence of mild to moderate mitral regurgitation or whether they were obtained after balloon valvotomy. In summary, linear regression analysis yields accurate estimates of MVA despite the theoretical superiority of nonlinear methods. On-line digital analysis of mitral gradient tracings may thus be useful at the time of diagnostic cardiac catheterization or balloon mitral valvotomy to assess the severity of mitral stenosis and the response to interventions.