The "trap" hypothesis of venous ulceration