Hepatic encephalopathy in patients with chronic liver dysfunction is believed to be caused by a failure of the liver to clear toxic products from the stomach. The exact toxins that cause hepatic encephalopathy have not been established, but ammonia may be involved. Many physicians determine ammonia levels to diagnose hepatic encephalopathy and as a guide to treatment. However, studies have shown that the correlation between serum ammonia levels and severity of hepatic encephalopathy is inconsistent. A recent study suggested that the partial pressure of ammonia may correlate more closely with the severity of hepatic encephalopathy than the total plasma ammonia level. Ong and associates evaluated the correlation between plasma ammonia levels and the severity of hepatic encephalopathy. They also determined the best of the four types of ammonia measurements for this correlation by comparing arterial total ammonia, venous total ammonia, arterial partial pressure of ammonia, and venous partial pressure of ammonia.

Consecutive patients who were admitted to a tertiary care center with the diagnosis of cirrhosis between September 1998 and December 1999 were enrolled in the study. The diagnosis of cirrhosis was established by biopsy or by signs of portal hypertension such as gastroesophageal varices, previous variceal bleeding, or ascites. Researchers collected clinical and laboratory data at the time of admission. The mental status of the patients was assessed using the West Haven Criteria for grading of mental status (see accompanying table). A diagnosis of hepatic encephalopathy was established when the patients' mental status was altered and other causes of mental status changes had been excluded. After diagnosis, fasting arterial and venous blood samples were obtained, and total ammonia and partial pressure of ammonia were determined.

The 121 patients who were enrolled in the study had the following West Haven Criteria grades: 30 patients (25 percent) had grade zero encephalopathy; 27 patients (22 percent) had grade 1; 23 patients (19 percent) had grade 2; 28 patients (23 percent) had grade 3; and 13 patients (11 percent) had grade 4. All four measurements of ammonia increased with the severity of hepatic encephalopathy. The arterial total ammonia level had the highest correlation, but it was not statistically significant. Other variables that had a correlation to the severity of hepatic encephalopathy included International Normalized Ratio (INR) values, serum creatinine levels, bilirubin levels, and lactulose use. A multivariable ordered logistic regression analysis revealed that only serum ammonia levels and INR values were independently associated with the severity of hepatic encephalopathy.

The authors conclude that venous total ammonia levels do correlate with severity of hepatic encephalopathy and should be adequate in evaluating patients with this condition. Total arterial ammonia levels and partial pressure of ammonia levels had similar correlation but did not prove to be better markers than venous total ammonia levels.