Noninvasive prediction of pulmonary artery pressure in patients with isolated ventricular septal defect

Management of patients with isolated ventricular septal defect (VSD) requires information regarding pulmonary artery pressure (PAP). The purpose of this study was to evaluate the individual predictive value of noninvasive methods for assessment of PAP and to determine if any combination of techniques significantly improved their predictive power. We reviewed the clinical history, electrocardiogram, and echocardiogram of 31 patients (age 1.9 +/- 1.73 years) who underwent catheterization for isolated VSD. Noninvasive data were compared for patients with mean PAP <20 mmHg (group 1) and those with mean PAP greater than or equal to 20 (group 2) at catheterization. Fourteen (45%) patients were in group 1 and 17 (55%) in group 2. Doppler estimation of VSD gradient, right ventricular hypertrophy by echocardiogram, interventricular septal orientation, and VSD size had predictive value for elevated mean PAP (p < 0.01). All patients (n = 6) with normal findings in all four variables had normal PAP. All patients (n = 12) with at least three of four abnormal findings had elevated PAP. Six patients in group 1 had at least one variable that incorrectly predicted high PAP, whereas 3 patients with normal findings on three of the four variables nevertheless had elevated PAP. No single noninvasive variable accurately predicted PAP in all cases. However, normal findings for all four significant variables did predict normal PAP and suggest that cardiac catheterization is unnecessary in that setting. However, any other combination of normal and abnormal findings for the four significant variables did not reliably predict PAP and such patients may require catheterization to directly measure PAP.