Effects of hydroxyethyl starch administration on renal function in critically ill patients

Background. The influence of hydroxyethyl starch (HES) solutions on renal function is controversial. We investigated the effect of HES administration on renal function in critically ill patients enrolled in a large multicentre observational European study. Methods. All adult patients admitted to the 198 participating intensive care units (ICUs) during a 15-day period were enrolled. Prospectively collected data included daily fluid administration, urine output, sequential organ failure assessment (SOFA) score, serum creatinine levels, and the need for renal replacement therapy (RRT) during the ICU stay. Results. Of 3147 patients, 1075 (34%) received HES. Patients who received HES were older [mean (SD): 62 (SD 17) vs 60 (18) years, P=0.022], more likely to be surgical admissions, had a higher incidence of haematological malignancy and heart failure, higher SAPS II [40.0 (17.0) vs 34.7 (16.9), P < 0.001] and SOFA [6.2 (3.7) vs 5.0 (3.9), P < 0.001] scores, and less likely to be receiving RRT (2 vs 4%, P < 0.001) than those who did not receive HES. The renal SOFA score increased significantly over the ICU stay independent of the type of fluid administered. Although more patients who received HES needed RRT than non-HES patients (11 vs 9%, P=0.006), HES administration was not associated with an increased risk for subsequent RRT in a multivariable analysis [odds ratio (OR): 0.417, 95% confidence interval (CI): 0.05-3.27, P=0.406]. Sepsis (OR: 2.03, 95% CI: 1.37-3.02, P < 0.001), cardiovascular failure (OR: 6.88, 95% CI: 4.49-10.56, P < 0.001), haematological cancer (OR: 2.83, 95% CI: 1.28-6.25, P=0.01), and baseline renal SOFA scores > 1 (P < 0.01 for renal SOFA 2, 3, and 4 with renal SOFA=0 as a reference) were all associated with a higher need for RRT. Conclusions. In this observational study, haematological cancer, the presence of sepsis, cardiovascular failure, and baseline renal function as assessed by the SOFA score were independent risk factors for the subsequent need for RRT in the ICU. The administration of HES had no influence on renal function or the need for RRT in the ICU.