Proximal isovelocity surface area (PISA) as a noninvasive method for the estimation of the shunt quantification in perimembranous ventricular septal defects

This study was designed to assess the reliability of the proximal isovelocity surface area (PISA) method for the estimation of shunt quantification in perimembranous ventricular septal defects (PVSD). The study group was composed of 30 patients (age 11 +/- 7 years, 13 female) with PVSD. The shunt flow Qp-Qs) and the ratio of the pulmonary flow to the systemic flow (Qp/Qs) were calculated by spectral Doppler and catheterization. The Qp-Qs, the defect area (DA), and the shunt volume (SV) were obtained by the PISA method. The PISA method estimated the DA (cm(2)/m(2)), the SV (cm(3)/m(2)), and the Qp-Qs (L/min/m(2)) to be equal to (2 x pi x R-2 x NL)/(V-max x Body surface area), DA x TVIshunt, and to SV X Heart rate, respectively (R is the distance of the maximal PISA from the first aliasing line to the left ventricular side of the defect, NL is the nyquist limit, and V-max and TVIshunt are the peak velocity and time-velocity integral of transdefect Doppler tracing obtained by continuous-wave Doppler). The PISA method (3.4 +/- 1.5 L/min/m(2)) underestimated the Qp-Qs according to spectral Doppler (r = 0.96 P < 0.001; mean difference -0.74 <plus/minus> 0.61 L/min/m(2); SEE = 0.11 L/min/m(2), P ( 0.001) and catheterization (r = 0.92, P < 0.001; mean difference -0.45 + 0.7 L/min/m(2): SEE = 0.13 L/min/m(2), P < 0.001). The correlations between the PISA findings Qp-Qs, DA, SV) and the catheterization Qp/Qs (r = 0.86 0.84, and 0.86; P < 0.001, respectively), or between these and the spectral Doppler Qp/Qs (r = 0.80 0.80 and 0.79; P < 0.001, respectively) were significant. The accuracies of the PISA findings in. identifying large defects were high (0.90 0.93 and 0.90 for cut-off values of Qp-Qs = 3.67 L/min/m(2) DA = 0.44 cm(2)/m(2) and SV = 43 cm(3)/m(2), respectively). As a result, the PISA method can be a simple and reliable alternative to the spectral Doppler method in the identification of large shunts in PVSD.