Fluid resuscitation: colloids may cause more patients to die than crystalloids.

Clinical bottom line (level 1a-)

  1. Patients who are critically ill, require fluid resuscitation and are given colloids are not clearly less likely to die than those given crystalloids, and they may well be more likely to.
Schierhout and Roberts: British Medical Journal 1998; 316: 961-964
Expires August 2003

The study

Systematic review of randomised controlled trials of
  • Patients: critically ill and requiring fluid resuscitation (excluding neonates)
  • Intervention: colloids compared with crystalloids
  • Outcome: death


    • Articles found in any using Cochrane Controlled Triasl Register, MEDLINE, EMBASE, BIDS Index to Scientific and Technical Proceedings, up to June 1997 (search terms: available from the authors on request ) and and searching of 1929 international journals and the proceedings of several international meetings on fluid therapy, checking reference lists of trials and review articles, contacting authors of all identified trials for information about other published and unpublished trials

    Selection criteria: as above
    Appraisal criteria: two independent blinded reviewers assessed eligibility
    Articles excluded if: crossover trials, resuscitation algorithms, intervention used for maintaining serum albumin concentration, haemodilution, fluid loading, reducing intracranial pressure

    Nineteen trials involving 1315 patients.
    There were no overall heterogeneity between the trials.

    The evidence

    Outcome Time to outcome CER RR
    (95% CI)    		 NNH
    (95% CI)
    death unknown 118/630
    (18.7%)    		 1.19
    (0.98 to 1.45)     		28
    (NNT = 263 to infinity;
    NNH = 12 to infinity)

    Comments

    1. No differences were noted for different types of injury requiring fluid resuscitation.
    2. Inclusion only of trials with adequate concealment failed to show any significant benefit or harm either 1.29 (95% CI, 0.94 to 1.77).
    3. There are not enough studies to indicate clearly whether using colloids is harmful.
    4. For now, the treatment of non-exsanguinating hypotension remains the judicious use of crystalloid while avoiding volume overload.

    Citation

    1. Schierhout G, and Roberts I: Fluid resuscitation with colloid or crystalloid solutions in critically ill patients: a systematic review of randomized trials. British Medical Journal 1998; 316: 961-964
    Search Terms: colloid in Cochrane
    Contributor: Catherine Clase, Chris Ball and Clare Wotton, April 2000
    Reviewer: Harold Szerlip

    Clinical Question.
    Patient requiring fluid resuscitation
    Intervention or Exposure colloids
    Comparison crystalloids
    Outcome death