Acute renal failure in intensive care units--causes, outcome, and prognostic factors of hospital mortality; a prospective, multicenter study. French Study Group on Acute Renal Failure.

Abstract

To assess the causes, the prognostic factors, and the outcome of patients with severe acute renal failure. Prospective, multicenter study. Twenty French multidisciplinary intensive care units (ICUs). All patients with severe acute renal failure were prospectively enrolled in the study for a 6-month period. Severe acute renal failure was defined by the following criteria: a) a serum creatinine concentration of > or = 3.5 mg/dL ( > or = 310 mumol/L) and/or a blood urea nitrogen concentration of > or = 100 mg/dL ( > or = 36 mmol/L); or b) an increase in blood urea nitrogen or serum creatinine concentration, such that the concentration is 100% above the baseline value in patients with previous chronic renal insufficiency (serum creatinine concentration of > 1.8 mg/dL [ > 150 mumol/L]), excluding those patients with a basal serum creatinine concentration of > 3.4 mg/dL ( > 300 mumol/L). None. Age, sex, previous health status and preexisting organ dysfunction, and type and origin of acute renal failure were recorded. The Simplified Acute Physiology Score, the Acute Physiology and Chronic Health Evaluation (APACHE II) score, and the number of Organ System Failures were calculated on ICU day 1 and at the time of inclusion in the study. Prognostic factors were determined by univariate methods and stepwise logistic regression analysis. There were 360 patients in the study; 217 patients were admitted to the study at the time of ICU admission and 143 patients were admitted to the study after ICU admission. Only 41% of these patients were in good health 3 months before ICU entry. The reason for admission was medical in 78% of cases. The type of acute renal failure was prerenal (n = 16), renal (n = 282), or postrenal (n = 17). Renal replacement therapy was used in 174 patients. Two hundred ten (58%) patients died during the hospital stay. Using stepwide logistic regression, seven variables were predictive of death. These variables were advanced age, altered previous health status, hospitalization before ICU admission, delayed occurrence of acute renal failure, sepsis, oliguria, and severity of illness as assessed at the time of study inclusion by Simplified Acute Physiology Score, APACHE II, or Organ System Failure. The hospital mortality rate of patients with severe acute renal failure in patients requiring intensive care remains high. In order to compare patient groups in further trials concerning acute renal failure, recorded characteristics of the population should include age, previous health status, disease characteristics (initial or delayed acute renal failure, oliguria, sepsis), and the severity of the illness as assessed by physiologic scoring systems recorded at the time of study inclusion.