PROGNOSTIC INDICATORS IN PATIENTS WITH CI RRHOSIS ADMITTED TO AN INTENSIVE-CARE UNIT

Citation
L. Castera et al., PROGNOSTIC INDICATORS IN PATIENTS WITH CI RRHOSIS ADMITTED TO AN INTENSIVE-CARE UNIT, Gastroenterologie clinique et biologique, 20(3), 1996, pp. 263-268
Citations number
20
Categorie Soggetti
Gastroenterology & Hepatology
ISSN journal
03998320
Volume
20
Issue
3
Year of publication
1996
Pages
263 - 268
Database
ISI
SICI code
0399-8320(1996)20:3<263:PIIPWC>2.0.ZU;2-U
Abstract
Objectives. - Assessment of prognosis in patients with cirrhosis admit ted to an Intensive Care Unit remains unsatisfactory. The aims of this retrospective study were to determine the survival rates of patients admitted to an Intensive Care Unit, and to identify and validate progn ostic indicators associated with a high mortality rate. Methods. - Two hundred and forty three patients with cirrhosis consecutively admitte d to the Intensive Care Unit were studied The main reasons for admissi on were upper gastrointestinal bleeding (n = 163), coma (n = 43), seps is (n = 18), and liver failure (n = 13). Patients were divided into tw o groups: group 1 (n = 121) to identify prognostic indicators associat ed with a high mortality rate, and group 2 (n = 122) to validate these indicators. Results. - Intensive Care Unit and one year survival rate s of patients with cirrhosis admitted for upper gastrointestinal hemor rhage were 76 and 50 % respectively. These rates were 40 and 8 % respe ctively for patients admitted for other reasons. In group 1, 4 predict ive factors found at admission were identified to have independent sig nificance by stepwise logistic regression: grade III or IV encephalopa thy, prothrombin index, serum creatinine, and hypoxemia. On the other hand, the presence of shock on admission was associated with a 100 % m ortality rate. Two prognostic indicators were defined: shock requiring the administration of vasoactive drugs, and the presence of 3 out of the 4 following predictive factors: grade III or IV encephalopathy, me chanical ventilation, prothrombin index < 30 %, and serum creatinine > 130 mu mol/L. In group 2, the presence of at least one prognostic ind icator at admission or during intensive care was associated with a 96 % mortality rate. These indicators were present in 69 % of patients wh o died. In 17 patients who died, but survived more than 24 hours in th e Intensive Care Unit, indicators were present an average of 6.0 +/- 5 .3 days before death. Conclusion. - Common prognostic indicators may a ccurately predict death in patients with cirrhosis admitted to an Inte nsive Care Unit. These indicators could be helpful in identifying pati ents who will not benefit from intensive care.