Influence of hydroxyethyl starch and gelatin versus crystalloids on renal function, fluid balance, and ICU length of stay in patients with severe sepsis

Introduction

In a smaller sample of patients with severe sepsis, resuscitation with the synthetic colloids hydroxyethyl starch (HES) as well as gelatin (GEL) increased the occurrence of acute kidney injury (AKI) in comparison with crystalloids (CRYS). We now performed the analysis in a large-scale sample with over 1,100 septic patients on our ICU.

Methods

A prospective controlled before-and-after study in patients with severe sepsis on a mixed ICU. AKI was defined by RIFLE criteria [1]. Statistical analysis was performed using SPSS 18.0.

Results

A total of 1,165 patients with severe sepsis were included. At baseline, the three groups had similar age, SAPS2 and SOFA scores and serum creatinine levels. Between January 2004 and January 2006, patients received fluid resuscitation with HES (median cumulative dose 81 ml/kg (IQR 38 to 157), n = 391), mainly as 6% HES 130/0.4 (in 75% of patients) or 10% HES 200/0.5. Between February 2006 and March 2008 patients received 4% GEL (40 ml/kg (IQR 18 to 71), n = 396), and between April 2008 through April 2010 patients received only CRYS (n = 387). AKI by any criteria (risk, injury or failure) was 34% after CRYS, 55% after HES, and 47% after GEL (P < 0.001 for HES or GEL vs. CRYS). Renal replacement therapy (RRT) was 28% after CRYS compared with 34% after HES (P = 0.04) or 39% after GEL (P = 0.002). Median cumulative fluid input during ICU stay was 659 ml/kg (IQR 269 to 1,250) after HES, 526 ml/kg (IQR 174 to 817) after GEL and 360 ml/kg (IQR 174 to 817) after CRYS (P < 0.001 HES vs. CRYS, P = 0.003 GEL vs. CRYS). Patients receiving synthetic colloids had a significantly longer median length of stay in the ICU (HES: 17 (IQR 8 to 29) days; GEL: 13 (IQR 6 to 24) days vs. CRYS: 11 (IQR 5 to 20) days (HES vs. CRYS P < 0.001, GEL vs. CRYS P = 0.001)). ICU mortality was 35% (HES), 32% (GEL), and 30% (CRYS, P = not significant).

Conclusions

Patients with severe sepsis have a higher risk to develop AKI if they receive fluid resuscitation with synthetic colloids (HES or gelatin). Interestingly, the need for RRT under fluid therapy with mainly 6% HES 130/0.4 was higher than in the VISEP study under therapy with 10% HES 200/0.5 (RRT: 31.1%, n = 261) [2].