We examined the cases of 31 patients over the age of 50 years undergoing operative closure of isolated ostium secundum atrial septal defect. The lesion had been diagnosed in all cases prior to cardiac catheterization. To assess the importance of pre-operative data on surgical outcome, the patients were first divided into three groups according to mean pulmonary artery pressure (PAP): less than 16 mmHg (Group A), 16-30 mmHg (Group B) and greater than 30 mmHg (Group C). Symptomatic improvement occurred in all groups but more patients in Group C, although symptomatically improved, remained short of breath and in atrial fibrillation than in Group A. Patients in Group A had a higher actual forced vital capacity expressed as a percentage of the predicted value (FVCa/FVCp) than patients in Group B or Group C (P less than 0.015). There was a good correlation between FVCa/FVCp and percentage oxygen saturation of the arterial blood (P less than 0.0009). This simple non-invasive investigation was therefore found to correlate with previously documented parameters, pulmonary artery pressure and percentage oxygen saturation of the arterial blood, affecting surgical outcome. Patients were also divided into groups according to FVCa/FVCp: less than 75% (Group 1), 50-75% (Group 2) and less than 50% (Group 3). Postoperative symptoms were more common in Group 3 than in Group 1. We conclude that respiratory function tests, as well as measurement of pulmonary artery pressures, are useful in predicting improvement following atrial septal repair.

We report a series of 21 consecutive patients undergoing combined multiple valve procedures and myocardial revascularisation between 1978 and 1988. There were 11 females and 10 males with a mean age of 58.5 (+/- 5.7) years. All patients were in NYHA Class 2 or more and 12 patients (57%) had angina. The mean left ventricular segment score was 7.9 (+/- 3.3). Five patients had undergone previous cardiac surgery. In all patients the aetiology of the valvular dysfunction was rheumatic. The first patient in the series was operated on using ischaemic arrest. The remaining 20 operations were performed using cardioplegia (2 crystalloid, 18 blood). A mean of 1.63 grafts per patient were inserted. There were 20 aortic valve replacements, 1 aortic valvotomy, 13 mitral valve replacements, 7 open mitral valvotomies, 1 mitral valve repair, and 1 tricuspid valve replacement. 1 patient had 3 valves replaced. Five deaths occurred in the series; all were due to low cardiac output and occurred prior to discharge from hospital. Follow-up ranged from 7 to 111 months (mean = 46 +/- 33). Three patients developed mitral paravalvular leaks, two of which were successfully repaired at 2 months and 2 years postoperatively. The third was asymptomatic. There were no late deaths and all survivors improved to NYHA Class 1 and had no angina. Early death was associated with increased perfusion time (p less than 0.01), the need for postoperative inotropic support (p less than 0.01) and high blood loss. No preoperative predictors of early death were identified. Multiple valve procedures and myocardial revascularisation carry a significant early mortality but are justified by the satisfactory long-term outcome.