July 24, 1995

Prognostic Stratification in Critically Ill Patients With Acute Renal Failure Requiring Dialysis

Author Affiliations

From the Renal Division, Department of Medicine, Brigham and Women's Hospital (Drs Chertow and Lazarus and Ms Munro), the Renal Section, Medical Service, Brockton—West Roxbury Department of Veterans Affairs Medical Center (Dr Chertow), and the Department of Health Care Policy, Harvard Medical School (Drs Christiansen and Cleary), Boston, Mass.

Arch Intern Med. 1995;155(14):1505-1511. doi:10.1001/archinte.1995.00430140075007

Background:  Despite the widespread availability of dialytic and intensive care unit technology, the probability of early mortality in critically ill persons with acute renal failure is distressingly high. Previous efforts to predict outcome in this population have been limited by small sample size and the absence of uniform exclusion criteria. Additionally, data obtained decades ago may not apply today owing to changes in case mix.

Methods:  The medical records of 132 consecutive patients in the intensive care unit with acute renal failure who required dialysis from 1991 through 1993 were evaluated by a blinded reviewer.

Results:  The overall in-hospital mortality rate was 70%. Twelve readily available historical, clinical, and laboratory variables were significantly associated with in-hospital mortality. Multivariate logistic regression analysis showed that mechanical ventilation, malignancy, and nonrespiratory organ system failure were independently associated with in-hospital mortality. Using a 95% positivity criterion, this model identified 24% of high-risk patients who died, without misclassification of any survivors. Of those who survived to hospital discharge, 33% were dialysis dependent and 28% were institutionalized long-term.

Conclusions:  Among critically ill patients, acute renal failure requiring dialysis is an ominous condition with a high risk of in-hospital mortality. This risk appears to depend largely on comorbid conditions, such as the need for mechanical ventilation and underlying malignancy. While this prognostic model requires prospective validation, it appears to identify a substantial fraction of patients for whom dialysis may be of limited or no benefit.(Arch Intern Med. 1995;155:1505-1511)