To the Editor.
—We have read the article by Rodzynek et al1 dealing with the diagnostic value of antithrombin III (AT III) determinations and the aminopyrine breath test in patients with liver diseases. We are pleased that these workers confirm the findings we reported earlier. In 1984, we reported, as did Rodzynek et al1 in 1986, that the decrease in AT III levels in patients with cirrhosis of the liver was greater than in patients with aggressive hepatitis and persistent hepatitis.2,3 The decrease in AT III levels was more marked in patients with cirrhosis of the liver and ascites than in patients with compensated cirrhosis of the liver. Also, AT III levels proved to be a more sensible and specific marker for cirrhosis of the liver than prothrombin time.4From these data, we conducted a survival study in cirrhotic patients, trying to correlate the decrease of AT