The New England Journal of Medicine NOVEMBER 15, 2012

Hydroxyethyl Starch or Saline for Fluid Resuscitation in Intensive Care

John A. Myburgh, Simon Finfer, Rinaldo Bellomo, et al.

Bottom Line

The CHEST trial demonstrated that in critically ill adults, resuscitation with 6% hydroxyethyl starch (130/0.4) compared to 0.9% saline did not reduce 90-day mortality but was associated with an increased requirement for renal-replacement therapy.

Key Findings

1. The primary outcome of 90-day mortality showed no significant difference between groups, occurring in 597 of 3,315 patients (18.0%) in the hydroxyethyl starch group compared to 566 of 3,336 patients (17.0%) in the saline group (relative risk 1.06; 95% CI 0.96-1.18; p=0.26).
2. Patients in the hydroxyethyl starch group had a significantly higher rate of renal-replacement therapy compared to the saline group (7.0% vs. 5.8%; relative risk 1.21; 95% CI 1.00-1.45; p=0.04).
3. Hydroxyethyl starch use was associated with an increased incidence of adverse events (5.3% vs. 2.8%; p<0.001), including pruritus and skin rash.
4. Fluid requirements were lower in the hydroxyethyl starch group, yet this volume sparing did not translate into improved clinical outcomes or survival.

Study Design

Design
RCT
Double-Blind
Sample
7,000
Patients
Duration
90 days
Median
Setting
Multicenter, Australia/New Zealand
Population Adult patients admitted to intensive care units requiring fluid resuscitation
Intervention 6% hydroxyethyl starch (130/0.4) in 0.9% sodium chloride
Comparator 0.9% sodium chloride (saline)
Outcome Death within 90 days of randomization

Study Limitations

The trial was a pragmatic study that allowed for clinician discretion in fluid administration, which may introduce variability in practice.
The study excluded patients with specific high-risk conditions such as severe burns, cardiac surgery, and liver transplantation, limiting the generalizability of these findings to those specific populations.
While statistically significant, the absolute difference in renal-replacement therapy use was relatively small, though clinically meaningful in the context of ICU management.

Clinical Significance

The CHEST trial provided definitive evidence that the widely used 'tetrastarch' hydroxyethyl starch (130/0.4) offers no survival benefit over standard crystalloids for ICU fluid resuscitation and carries a higher risk of renal injury requiring dialysis, leading to a significant shift in clinical practice and international guidelines against its use in critically ill patients.

Historical Context

Prior to the CHEST trial, hydroxyethyl starch solutions were frequently employed as a colloid to achieve hemodynamic goals with smaller fluid volumes compared to crystalloids. Despite persistent concerns regarding nephrotoxicity and coagulopathy, modern third-generation starches (130/0.4) were initially thought to be safer; this trial was instrumental in rigorously testing that hypothesis and ultimately refuting the perceived clinical benefit of these colloids in the ICU setting.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

Physiologically, what is the theoretical advantage of using a colloid like hydroxyethyl starch (HES) over a crystalloid like 0.9% saline for volume resuscitation, and what mechanism likely explains the increased need for renal-replacement therapy observed in the CHEST trial?

Key Response

Colloids exert higher oncotic pressure than crystalloids, theoretically remaining in the intravascular space longer and requiring less volume to achieve hemodynamic targets (volume sparing). However, HES molecules are filtered by the glomerulus and can be taken up by proximal tubular cells via pinocytosis, leading to osmotic nephrosis and cellular swelling, which causes acute kidney injury.

Resident
Resident

In a patient presenting with severe sepsis and early signs of acute kidney injury (AKI), how should the findings of the CHEST trial influence your choice of resuscitation fluid, and how do these findings align with current Surviving Sepsis Campaign recommendations?

Key Response

The CHEST trial found that HES significantly increased the need for renal-replacement therapy (RRT) without providing a mortality benefit. Consequently, the Surviving Sepsis Campaign guidelines now issue a strong recommendation against using HES for fluid resuscitation in patients with sepsis, prioritizing crystalloids as the first-line choice.

Fellow
Fellow

The CHEST trial observed a 'volume-sparing effect' with HES (lower volumes required in the first 4 days) but failed to show a difference in 90-day mortality. Discuss the clinical significance of volume-sparing in light of the 'safety signal' for RRT and how this affects your management of patients with refractory shock.

Key Response

While HES achieved resuscitation with slightly less volume, this benefit is considered clinically negligible compared to the 21% relative increase in RRT requirement. In refractory shock, rather than switching to HES for volume-sparing, fellows should focus on early vasopressor initiation and consideration of other colloids like albumin, which do not share the same nephrotoxic profile as HES.

Attending
Attending

Reflecting on the 'Colloid vs. Crystalloid' debate, how did the CHEST trial contribute to the shift in ICU practice from evaluating fluid efficacy based on hemodynamic surrogate markers to evaluating safety based on patient-centered outcomes like RRT and long-term mortality?

Key Response

For decades, fluids were judged by their ability to raise CVP or MAP quickly. CHEST (alongside VISEP and 6S) shifted the paradigm by demonstrating that even if a fluid is hemodynamically 'efficient,' its side-effect profile (nephrotoxicity) can result in significant morbidity, effectively ending the routine use of synthetic colloids in the ICU.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The CHEST trial used 90-day mortality as the primary endpoint. Critique the use of this endpoint versus a composite renal endpoint in a fluid trial, particularly considering the potential for 'competing risk of death' to bias the assessment of renal-replacement therapy requirements.

Key Response

90-day mortality is a robust, patient-centered endpoint but may be too distal to capture the specific harms of a fluid. Conversely, RRT as an endpoint is susceptible to survivor bias; patients in the saline group who died early might have required RRT had they lived, potentially inflating the relative safety of saline or masking the true nephrotoxicity of HES. A competing risk analysis (e.g., Fine-Gray model) is essential for such interpretations.

Journal Editor
Journal Editor

The CHEST trial was a large-scale, pragmatic, double-blind study. As a reviewer, how would you evaluate the threat to internal validity posed by the use of non-study fluids (contamination) during the pre-randomization and maintenance phases of the trial?

Key Response

Contamination (where patients receive the opposite or other fluids) generally biases results toward the null, making it harder to find a difference. In CHEST, many patients received crystalloids before randomization. However, the fact that a significant difference in RRT was still detected despite this potential dilution of effect actually strengthens the conclusion that HES is harmful.

Guideline Committee
Guideline Committee

When comparing the CHEST trial results to the earlier VISEP and 6S trials, how does the larger, more heterogeneous population in CHEST affect the 'Strength of Recommendation' for avoiding HES in non-septic critically ill patients?

Key Response

While VISEP and 6S focused heavily on sepsis, CHEST included a broad range of ICU patients (trauma, post-op, etc.). By showing increased RRT risk even in this more heterogeneous cohort without any mortality benefit, CHEST allows guideline committees to extend the 'Strong Recommendation' against HES beyond just sepsis to almost all critically ill adults.

Clinical Landscape

Noteworthy Related Trials

2008

VISEP Trial

n = 537 · JAMA

Tested

Hydroxyethyl starch (HES) 10%

Population

Patients with severe sepsis or septic shock

Comparator

Modified Ringer lactate

Endpoint

Organ failure assessment (SOFA score)

Key result: HES resuscitation did not improve organ failure but was associated with increased risk of acute kidney injury and need for renal replacement therapy.
2012

6S Trial

n = 798 · NEJM

Tested

Hydroxyethyl starch (tetrastarch) 130/0.4

Population

Patients with severe sepsis

Comparator

Ringer's acetate

Endpoint

Composite of death or end-stage kidney failure at 90 days

Key result: Patients randomized to HES had a higher rate of death and were more likely to require renal-replacement therapy compared to those receiving crystalloids.
2015

SPLIT Trial

n = 2,278 · JAMA

Tested

Buffered crystalloids

Population

ICU patients requiring fluid resuscitation

Comparator

0.9% saline

Endpoint

Acute kidney injury defined by AKIN criteria

Key result: There was no significant difference in the incidence of acute kidney injury or other clinical outcomes between buffered crystalloids and saline.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis