Early goal-directed therapy in the treatment of severe sepsis and septic shock

Rivers. NEJM 2001; 345:1368-1377. doi:10.1056/NEJMoa010307

Clinical Question

  • In adults with severe sepsis or septic shock, does the use of early goal-directed therapy (EGDT) reduce the mortality?


  • Randomised controlled trial
  • Computer-generated blocks of 2 to 8
  • Concealed allocation
  • No blinding of clinicians, study personnel or patients
  • Power calculation: sample size of 260 required with 80% power and 0.05 alpha error rate to detect absolute reduction in mortality of 15% (expected baseline not specified)
  • Intention-to-treat analysis


  • Single academic tertiary hospital in Detroit, USA
  • March 1997 to March 2000


  • Inclusion: Severe sepsis or septic shock
    • ≥ 2 SIRS features, and either
    • systolic BP ≤ 90 mmHg after crystalloid fluid of 20 to 30 ml/kg over 30 minutes, or
    • blood lactate ≥ 4 mmol/l
  • Exclusion: age under 18 years; pregnancy; other acute neurological, cardiac, gastrointestinal or respiratory pathology; contraindications to central catheterisation; uncured cancer; immunosupression; limitations defined for active care
  • 268 randomised with 236 completing study


  • Early goal-directed therapy for 6 hours
    • Remained in ED for 6 hours and protocolised therapy managed by an emergency physician, two residents and three nurses
    • Oxygen ± intubation and mechanical ventilation
    • Central venous and arterial catheterisation
    • Central venous catheter (Edwards Lifesciences) capable of measuring continuous ScvO
    • 500 ml fluid boluses until CVP 8–12 mmHg
    • Vasopressors or vasodilators until MAP 65–90 mmHg
    • If ScvO2 < 70%
      • Transfused red cells until haematocrit ≥ 30%
      • If still ScvO2 < 70%
        • Dobutamine 2.5–20 µg/kg/min until ScvO2 > 70%
        • Dobutamine limited or reduced if MAP < 65 mmHg or pulse > 120 /min


  • Usual care for 6 hours without protocol
    • Treatment at the discretion of the clinician and patient admitted to ICU as soon as possible after consult and lines inserted
    • Oxygen ± intubation and mechanical ventilation
    • Central venous and arterial catheterisation
    • Targets defined but without treatment algorithm:
      • CVP ≥ 8–12 mmHg
      • MAP ≥ 65 mmHg
      • Urine output ≥ 0.5 ml/kg/hr


  • Primary outcome: in-hospital mortality was statistically and clinically significantly better in the EGDT intervention group compared to the usual care control group.
    • 30.5% vs 46.5% p = 0.009
  • Secondary outcome: subgroup analyses all favoured EDGT
    • severity of sepsis
    • 28-day and 60-day mortality
    • cause of in-hospital death (sudden cardiovascular collapse and multi-organ failure)
  • Tertiary data collection:
    • The EGDT group received more fluid over the early 6-hour period but there was no difference in total fluid over 72 hours
    • More red-cell transfusions were given to the EGDT group
    • Early inotropic use was more prevalent in the EGDT group (not statistically different over complete 72-hour period)
    • Vasopressor use was more prevalent in the usual care group
    • Intubation and ventilation was more prevalent in the usual care group
Primary Outcome
Measure EGDT Usual Care ARR 95% CI p NNT
In-hospital mortality 29.2% 44.4% 15.13% 3.62 to 26.64% 0.0150 7
ARR = absolute risk reduction; CI = confidence interval; p = p value; NNT = number needed-to-treat
Fishers’ exact test performed on reported in-hopsital mortality. This is different to Kaplan-Meier survival estimates reported in the orginal paper.

Authors’ Conclusions

  • “goal-directed therapy provided at the earliest stages of severe sepsis and septic shock… has significant short-term and long-term benefits.”


  • Adequate trial design for the clinical question
  • 90% completed the study – a drop-out / lost-follow-up rate that is well below the primary outcome rate


  • Single centre study
  • Intervention was a bundle of care, and individual components cannot be investigated separately
  • Usual care control group had a higher than average mortality rate
  • In-hospital mortality as an outcome measure has its limitations, as some patients will go to institutions or home to die soon after
  • Lead author is a paid consultant for manufacturer of ScvO2-measuring central catheter, and has registered “early goal-directed therapy” and “EGDT” as his trademarks!

The Bottom Line

  • Until ProCESS, ARISE and ProMISe, Rivers’ trial represented the best evidence for EGDT and demonstrated a strong benefit.
  • Concerns and criticism by clinicians has led to modified but widespread adoption of the ‘bundle’ of care in severe sepsis and septic shock.
  • Until individual elements are tested, the specific goals are not unreasonable goals to target, although the algorithm is perhaps less robust.

External Links


Summary author: @DuncanChambler
Summary date: 3 October 2014
Peer-review editor: @stevemathieu75


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.