Abstract
Eleven plasma variables have been studied in 18 elderly patients with cardiac oedema treated with bendrofluazide, frusemide and bumetanide, using a fully balanced, cross-over trial design. Bumetanide caused the least disturbance of plasma potassium and bendrofluazide the greatest. Frusemide produced an intermediate effect. The difference between the changes in plasma potassium produced by bumetanide and bendrofluazide was significant (P less than 0.02).