End-stage renal disease and outcome in a surgical intensive care unit

End-stage renal disease (ESRD) is associated with an increased propensity for critical illness, but whether ESRD is independently associated with a greater risk of death after major surgical procedures is unclear.

This was a retrospective analysis of prospectively collected data from all adult (>18 years) patients admitted to a 50-bed surgical intensive care unit (ICU) between January 2004 and January 2009. ESRD was defined as the need for chronic peritoneal dialysis or hemodialysis for at least 6 weeks prior to ICU admission. We used multivariable logistic regression analysis and propensity-score matching to adjust for possible confounders.

In total, 12,938 adult patients were admitted during the study period; 199 patients had ESRD at ICU admission, giving a prevalence of 1.5%. Patients with ESRD were more likely to be male (72.9% versus 63.0%, P = 0.004) and had higher severity scores, a higher incidence of diabetes mellitus and cirrhosis, and a lower incidence of cancer at ICU admission than those without ESRD. Patients with ESRD were more likely to have any type of organ failure at ICU admission and during the ICU stay. Patients with ESRD had higher ICU and hospital mortality rates (23.1% and 31.2% versus 5.5% and 10.0%, respectively, P <0.001 pairwise) and longer ICU length of stay (2 (1 to 7) versus 1 (1 to 3) days, P <0.001). In multivariable logistic regression analysis, ESRD was independently associated with a greater risk of in-hospital death (odds ratio = 3.84, 95% confidence interval 2.68 to 5.5, P <0.001). In 199 pairs of patients, hematologic and hepatic failures were more prevalent, ICU and hospital mortality rates were higher (23.1% versus 15.1% and 31.2% versus 19.1%, P <0.05 pairwise), and ICU length of stay was longer (2 (1 to 7) versus 1 (1 to 7) days, P <0.001) in patients with ESRD.

In this large cohort of surgical ICU patients, presence of ESRD at ICU admission was associated with greater morbidity and mortality and independently associated with a greater risk of in-hospital death. Our data can be useful in preoperative risk stratification.