Abstract
Congenital pulmonary arteriovenous fistulae, provided that they are not present in very large numbers, are best managed surgically, and there is no effective alternative management. Where pulmonary arteriovenous fistulae develop as a consequence of pulmonary hypertension it is rational to resect the fistulae if the cause of pulmonary hypertension - for example, mitral stenosis - is correctable. Pulmonary arteriovenous fistula in the presence of unexplained or uncorrectable pulmonary hypertension may be the safety valve on which life depends and should, therefore, not be resected. An example is reported of pulmonary arteriovenous fistula associated with bilharzial pulmonary hypertension in which resection of the fistula resulted in death.