Crystalloids or Colloids in Hypovolemic Shock


To test whether the use of colloids or crystalloids for fluid resuscitation is more likely to reduce mortality in patients admitted to the ICU with acute hypovolemic shock.


Two types of solution are used for expansion of fluid volume in patients in the ICU: crystalloids (e.g. isotonic or hypertonic saline or buffered Ringer lactate, acetate, or maleate) and colloids (e.g. gelatins, albumins, dextrans and hydroxyethyl starches). It is unclear, however, whether colloids or crystalloids are the better option for hypovolaemic patients in the ICU setting. The study reported here, ‘Colloids versus crystalloids for the resuscitation of the critically Ill (CRISTAL)’, compares patient outcomes after use of any agent from the class of colloids with any from the class of crystalloids.

Study design

An international, multicentre, randomized, open-label, parallel-group trial.

Patients and methods

The study ran from February 2003 to November 2012 in 57 ICUs in France, Belgium, Algeria, Tunisia and Canada. Eligible patients requiring fluid resuscitation for acute hypovolaemia were given either crystalloids (isotonic or hypertonic saline or Ringer lactate solution) or colloids (gelatins, dextrans, hydroxyethyl starches, or 4% or 20% of albumin) for all fluid interventions for the remainder of their stay in the ICU (other than crystalloids for routine fluid maintenance or albumin to counteract hypoalbuminaemia). The primary outcome measure was death within 28 days. Secondary outcomes were 90-day mortality, days alive and not receiving renal replacement therapy, mechanical ventilation or vasopressor therapy, days without organ system failure (i.e. SOFA score <6) and days not in the ICU or hospital.


Of 2857 patients entering the study, 1414 received colloids and 1443 received crystalloids. Between-group baseline characteristics were similar; most patients had severe sepsis at ICU admission. There was no significant between-group difference for the primary endpoint of death rates within 28 days: there were 359 deaths (25.4%) in the colloid-treated group vs. 390 deaths (27.0%) in the crystalloid-treated group (RR, 0.96; 95% CI, 0.88 to 1.04; p=0.26). There were, however significant differences in favour of colloids for several of the secondary endpoints. At 90 days there were 434 deaths (30.7%) in the colloid-treated group vs. 493 deaths (34.2%) in the crystalloid-treated group (RR, 0.92; 95% CI, 0.86 to 0.99; p=0.03). There was less requirement for mechanical ventilation in those given colloids vs. those given crystalloids both at 7 days (mean days without mechanical ventilation: 2.1 vs. 1.8 days; mean difference, 0.30; 95% CI, 0.09 to 0.48 days; p=0.01), and at 28 days (mean: 14.6 vs. 13.5 days; mean difference, 1.10; 95% CI, 0.14 to 2.06 days; p=0.01). In addition, there were more days without vasopressor therapy in the colloid-treated group by 7 days (mean: 5.0 vs. 4.7 days; mean difference, 0.30; 95% CI, −0.03 to 0.50 days; p=0.04) and by 28 days (mean: 16.2 vs. 15.2 days; mean difference, 1.04; 95% CI, −0.04 to 2.10 days; p=0.03). There was no significant between-group difference for the number of patients requiring renal replacement therapy, days without renal replacement therapy, patients with organ failure or days alive without organ failure or for ICU- or hospital-free days. There were also no significant between-group differences when the data were stratified by the cause of the initial hypovolaemic shock (i.e. trauma, sepsis or other causes).


Among ICU patients with hypovolaemia, there was no significant difference in 28-day mortality between patients given colloids and those given crystalloids, though colloids appeared to reduce the requirement for mechanical ventilation and vasopressor therapy while the patients were in the ICU. There were fewer deaths at 90 days in the patients treated with colloids, but the authors suggest that further studies are required to substantiate this finding.


This large study completed the series of big fluid trials in 2013 after CHRYSTMAS, 6S and CHEST being published in 2012. The study has some strengths: not only was it a pragmatic, international, randomized trial performed in 2 parallel groups of adult critically ill patients requiring fluid resuscitation for acute hypovolemia, including almost all consecutive patients at a very early stage of shock (because of a waiver of informed consent); it was also the only trial among the 4 where patients, in order to be eligible, had to have received no prior fluids for resuscitation during their ICU stay. However, there are many limitations:

  1. First, although a lot of patients were included, the study was initially aimed to enrol 3010 patients but was stopped at the 6th interim analysis due to the absence of difference in mortality at 28 days, with a final recruitment of 2857 patients, maybe the results could have been different when the proper power analysis was taken into account.
  2. Second, the patients were randomised to receive any crystalloid (balanced or unbalanced, isotonic or hypertonic) or any type of colloid (gelatines, HES, dextrans, but also hypotonic albumin 4% or hypertonic albumin 20%) as the resuscitation fluid (not the maintenance fluid), so all patients received a mix of colloids and crystalloids which resulted in a very heterogenous patient population.
  3. Third, the study was open-label and not-blinded which could have introduced a bias.
  4. Fourth, the study took 9 years to complete, and within this time-period a lot of practices may have changed impacting the results.
  5. Fifth, acute hypovolaemia was defined by the combination of 1) hypotension; 2) evidence for low filling pressures and low cardiac index as assessed either invasively or noninvasively; 3) signs of tissue hypoperfusion or hypoxia, however fluids were not really guided per protocol but rather at the discretion of the attending physician.
  6. Sixth, the median amount of study fluid in 1st 7 days was only 2L in colloid vs 3L in crystalloid and there were no differences in blood pressure, urine output, weight, fluid balance on the first day.

The Take home message

As of today, despite the recent large trials we still don’t know which type of fluid is the best in each phase of shock in each type of patient. We do know that the best fluid is probably the one that has not been given and fluid administration must be seen as any other drug where the risks and benefits are weighed; fluids must be guided per protocol and fluid responsiveness must be assessed upfront and above all fluid overload must be avoided and de-resuscitation must be considered. Taken this into account it seems that a dynamic approach starting with crystalloid maintenance and resuscitation fluids, followed by colloids (sepsis and burns excluded) or albumin at a later stage is our best option. So it’s all about the choice, the timing, the dose, and the speed of administration. Unfortunately one size does not fit all…


Annane D, Siami S, Jaber S et al. Effects of fluid resuscitation with colloids vs crystalloids on mortality in critically ill patients presenting with hypovolemic shock: the CRISTAL randomized trial. JAMA 2013;310:1809–1817. Trial registration number: NCT00318942.

Leave a comment

You are commenting as guest.

Social Media

  like us on Facebook

  follow us on Twitter

  join Discussion group

  join us on Linkedin

  newsletter sign up

  post on the blog