Identifying critically ill patients at high risk for developing acute renal failure: a pilot study

Kidney Int. 2005 Nov;68(5):2274-80. doi: 10.1111/j.1523-1755.2005.00686.x.


Background: Acute renal failure (ARF) occurs commonly in the intensive care unit (ICU), but predicting which patients will develop ARF is difficult. We set out to determine which risk factors would predict the development of ARF in critically ill patients who are admitted to the ICU without ARF.

Methods: From August 2002 to April 2003, we enrolled medical-surgical ICU admissions into a cohort using a sampling tool based on their risk factor (RF) profile. The risk factors we identified were separated into 3 categories: chronic major, chronic minor, and acute RFs. Combinations of these RFs were used to create a sampling tool and identify patients to enroll into our cohort. Patients with end-stage renal disease and ARF upon admission to the ICU were excluded.

Results: We enrolled 194 patients over a 14-month period. The mean age of the cohort was 64.6 +/- 14.7 years. The percentage of Caucasians, African Americans, and Hispanics was 40.7%, 50.5%, and 3.6%, respectively. In a univariate analysis of the entire cohort, increasing APACHE II quartile, increased A-a gradient, presence of systemic inflammatory response syndrome (SIRS), decreased levels of serum albumin, and presence of active cancer predicted ARF. In a multiple logistic regression analysis, decreased serum albumin (high levels of serum albumin were protective), increased A-a gradient, and cancer were associated with development of ARF (OR 2.17, 1.04, and 2.86, respectively).

Conclusion: Decreased levels of serum albumin concentration, increased A-a gradient, and presence of active cancer predict which patients who are admitted to the ICU will develop ARF.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Acute Kidney Injury / diagnosis*
  • Acute Kidney Injury / epidemiology*
  • Aged
  • Biomarkers
  • Creatinine / blood
  • Critical Illness / epidemiology*
  • Female
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Pilot Projects
  • Predictive Value of Tests
  • Risk Factors
  • Serum Albumin / metabolism


  • Biomarkers
  • Serum Albumin
  • Creatinine