Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Which fluid (colloids or crystalloids) is better in initial resuscitation of severe sepsis?

Three Part Question

In [adult patients with severe sepsis], should [colloids or crystalloids be used for initial resuscitation] to [improve survival].

Clinical Scenario

A 67-year-old woman presents to the Emergency Department (ED) with a 3-day history of a dysuria, nausea, vomiting and fever with rigours. She is confused and looks pale. Her respiratory rate is 40/min, heart rate is 120/min, blood pressure is 80/40 mm Hg and temperature is 38.9°C. You diagnose severe sepsis secondary to a urinary tract infection. You wonder whether crystalloids or colloids are best in the initial fluid resuscitation to improve her survival.

Search Strategy

Medline 1960 – 09/2008 using NLH 2.0 Advanced Search

({[exp sepsis] AND [early.ti.ab OR initial.ti.ab AND management.ti.ab OR resuscitation.ti.ab]} AND {[crystalloid$.af OR OR hartmann’ OR] AND [colloid$.af OR OR OR OR]}) AND Limit to humans, adults, English language and studies where survival is primary end point.

In addition reference lists were reviewed to identify additional papers not found in the computerised search.

Repeated March 2012

Search Outcome

12 papers were found of which 4 were irrelevant. One paper was revised in a later paper and one of insufficient quality for inclusion.
The 6 remaining papers and 3 more relevant papers found on reviewing reference lists are shown in the table in chronological order.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Schierhout G and Roberts I
A variety of critically ill patient groups were included in this review of 26 trials. Mortality data for 1315 patients from 19 trials. Systematic review of randomized trialsMortality from all causes at end of follow up periodColloid was associated with an increased absolute risk of mortality of 4% (95% CI 0-8%) when compared with crystalloid. Pooled relative risk of mortality for colloid was 1.29 (95% CI 0.94-1.77)Incomparable mixed patient groups including sepsis. Different fluid regimens between trials. Length of follow up varied from 1 day to 5 years. Only 19 of the 26 trials reported on mortality. There were concerns regarding the methodological quality of several of the included trials. Statistically not significant as CI crosses 1.
Choi P, Yip G et al
Adult patients (814) requiring fluid resuscitation in 17 trials. Systematic review of randomized trialsMortality No apparent difference between isotonic crystalloid and colloid resuscitationPower of aggregated data was insufficient. Methodological limitations of primary studies, including blinding, were noted. Only 3 included critically ill non-trauma patients including sepsis.
Rivers E et al
263 patients with severe sepsis/ septic shock presenting to an urban ED. Prospective randomized trialIn hospital mortality within 60 days Higher mortality in standard therapy (ST) group than in the Early Goal Directed Therapy (EGDT) group (P=0.009) ST group received fluids at clinicians’ discretion. EGDT group received 500ml boluses of crystalloid every 30 minutes +/- blood transfusions. Possibility of subversion of randomization in small blocks of pre-randomized numbers. ST patients stayed significantly shorter duration in ED; where EGDT patients at least for 6 hours. 10.25% (27) didn’t complete initial 6hour study period. ST group received some form of goal directed therapy. Supported by industry.
Finfer S et al (The SAFE Study investigators)
Australia & New Zealand
7000 adults (>18 years) in ITU requiring fluid treatment; 1218 were with severe sepsis. 4% Albumin vs normal saline Prospective double-blind randomized controlled trailDeath from any cause within 28 days after randomly included in the trial. All patientsRelative risk of death for albumin 0.99 (95% CI 0.91-1.09 P=0.87) when compared with saline3.8% (267) didn’t complete the trial. Under powered especially for subgroup analysis. Other fluids been administered due to error & clinicians’ preference. Greater volume of blood been transfused for albumin group than saline group. Possibility that concurrent interventions could have influenced results. Funded by CSL; 2 authors own shares in CSL.
Patients with severe sepsisRelative risk of death for albumin 0.87 (95% CI 0.74-1.02 P=0.09) when compared with saline
Vincent J and Gerlach H
A systematic review, giving graded recommendations (A to E, A being the highest) using modified Delphi methodology.Should colloids be used in preference to crystalloids in initial resuscitation from septic shock?Recommendation – Grade C (Uncertain) No evidence-based support for one type of fluid over anotherNone of the meta-analyses specifically focused on patients with severe sepsis/ septic shock.
McIntyre LA et al
496 patients with severe sepsis admitted to ICU. A retrospective multicentre cohort studyHospital Mortality; ICU MortalityNo association between hospital or ICU mortality and type of fluid (crystalloids vs crystalloids + colloids) administered in the first six hours after diagnosis of severe sepsis was identified.No evidence of sample size calculation. Potential for selection bias, information bias, confounding & residual confounding. Excluded patients’ (1643) charts were not screened for errors. Limited to ITU patients.
Dellinger RP et al
Surviving Sepsis Campaign:
Revision of 2004 guidelines based on an updated search into 2007 using GRADE system, a structured system for rating quality of evidence & grading strength of recommendations. High (A) to very low (D) quality of evidence. Strong recommendation (1) & weak (2).Which fluid/s to use in resuscitation?Recommends fluid resuscitation using crystalloids or colloids (1B)Some authors have financial and/or academic competing interests in industry
McIntyre et al
Hypotensive adult patients with at least 2 SIRs criteria and suspected or confirmed sepsis presenting to 3 centres in Canada and 1 in New Zealand.Prospective multicentre randomised pilot trial comparing saline with pentastarch fluids for resuscitation.ITU mortality6/19 patients in the saline group died while on ITU compared with 6/21 patients in the pentastarch group. (P = .84).Small study, designed as a pilot study. Study stopped early due to published data suggesting an increased risk of renal failure in septic patients treated with pentastarch fluids.
28 day mortality6/19 patients in the saline group died within 28 days compared with 9/21 patients in the pentastarch group. P = .46.
12 trials involving total of 1062 patients.Systematic review of randomised trials comparing the use of hydroxyethyl starch fluids with other fluids for resuscitation of patients with sepsis. Renal failureFluids based on hydroxyethyl starch increase the risk of acute renal failure among patients with sepsis.Meta-analysis not performed.
MortalityFluids based on hydroxyethyl starch for fluid resuscitation in patients with sepsis may reduce the probability of survival.


Fluid resuscitation is an important part of the initial management of patients with severe sepsis. Studies to date have not shown that either crystalloids or colloids are superior during the initial resuscitation of patients with severe sepsis. There are also other factors that need to be addressed regarding fluid choice. The choice of crystalloid is commonly between 0.9% sodium chloride and the more balanced salt solution Hartmann's. There are several different colloid solutions (starch, gelatine, dextran and albumin). Colloids may also be suspended in 0.9% sodium chloride or in a balanced salt solution. Furthermore, administration of hydroxyethyl starch may increase the risk of acute renal failure in patients with sepsis (Schortgen et al). The best choice from these options requires further study. The British consensus guidelines on intravenous fluid therapy for adult surgical patients support the use of balanced salt solutions (Powell-Tuck et al). A well designed and adequately powered randomised controlled trial with clinically important endpoints is required to address this question.

Clinical Bottom Line

Both crystalloids and colloids can be used in the initial resuscitation of patients with severe sepsis. The choice of fluid is probably less important than the quantity given. The current most Surviving Sepsis Campaign guidelines recommend giving fluid challenges of 1000 ml of crystalloids or 300–500 ml of colloids over 30 min to achieve a target CVP of 8 mm Hg or more (12 mm Hg if mechanically ventilated).

Level of Evidence

Level 1 - Recent well-done systematic review was considered or a study of high quality is available.


  1. Schierhout G and Roberts I Fluid resuscitation with colloid or crystalloid solutions in critically ill patients: a systematic review of randomised trials BMJ 1998; 316; 961-964
  2. Choi P, Yip G et al Crystalloids vs. colloids in fluid resuscitation: a systematic review Critical Care Medicine 1999; 27(1); 200-210
  3. Rivers E et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. The New England Journal of Medicine. 2001; 345:1368-1377
  4. Finfer S et al (The SAFE Study Investigators). A comparison of albumin and saline for fluid resuscitation in the intensive care unit. The New England Journal of Medicine 2004; 350; 2247-56.
  5. Vincent JL, Gerlach H. Fluid resuscitation in severe sepsis and septic shock: An evidence based review. Critical Care Medicine 2004; 32(11); 451-454
  6. McIntyre LA et al. Resuscitating patients with early severe sepsis: a Canadian multicentre observational study. Canadian Journal of Anaesthesia 2007; 54(10); 790-798.
  7. Dellinger RP et al. Surviving sepsis campaign: International guidelines for management of severe sepsis and septic shock 2008. Critical Care Medicine 2008; 36(1); 296-327
  8. Schortgen F, Lacherade JC, Bruneel F, et al. Effects of hydroxyethyl starch and gelatin on renal function in severe sepsis: A multi-centre randomized study. Lancet 2001; 357: 911-916.
  9. Powell-Tuck J, et al. British consensus guidelines on intravenous fluid therapy for adult surgical patients. GIFTASUP 2008
  10. McIntyre L, Fergusson D, Cook DJ, et al. Fluid resuscitation in the management of early septic shock (FINESS): a randomised controlled feasibility trial Can J Anaesth 2008; 55 (12) p819-826
  11. Wiedermann C Systematic review of randomised clinical trials on the use of hydroxyethyl starch for fluid management in sepsis BMC Emergency Medicine 2008; 8:1