Early NA in sepsis

Early vs delayed administration of norepinephrine in patients with septic shock. Bai et al. Critical Care 2014, 18:532

Appraisal by Dr. Peter Tsim. 


  • Study investigating delayed NA use following onset of septic shock on mortality
  • Retrospective study
  • End-point was 28 days mortality (target in sepsis)
  • Timing of vasopressors more important than choice of agent
  • Based on Surviving Sepsis guidelines:
    • 30ml/kg crystalloid, culture, lactate & abx within 3 hours
    • Use of NA to maintain MAP >65 within 6 hours
  • Fluid resuscitation preferred to inotrope use
  • Using too much fluid can increase extravascular oedema, reduce pulmonary function à reduce oxygenation of tissues
  • Question of when to start inotropes – too late can cause irreversible damage due to poor perfusion, too early can adversely affect organ microcirculatory blood flow



  • Selection of 213 patients from 2 general surgical ITUs of Jinling Hospital in China
  • Retrospective notes review
  • Sepsis Six applied as soon as septic shock identified (SIRS with infection)
  • Vasopressors started when Drs felt hypotension not responding to fluid resuscitation, i.e. hypotension persisting despite 30ml/kg, or hypotension resolving for <1 hour with 30ml/kg
  • All other causes of shock excluded (cardiogenic, haemorrhage), also patients with persisting hypotension >1 hour prior to admission



  • Variety of statistical analysis used
    • Continuous variables, normal distribution – Student T-test
    • Continuous variables, non-normal distribution – Mann-Whitney U test
    • Categorical variables – Chi-Squared
  • (Assuming normal distribution, 68% lie within +/-1 (SD), 95% lie withint +/-2 SD)
  • Overall 28 day mortality = 37.6%
  • No significant different between survivors and non-survivors with respect to age, gender, positive cultures, source of infection
  • Statistically significant results (survivor vs non-survivor)
    • Initial lactate 4.3 +/- 1.4 vs 5.3 +/- 1.5 (p<0.001)
    • APACHE II score 27.2 +/- 3.9 vs 30.4 +/- 3.9 (p<0.001)
    • 24 hour NA requirement (mg) 29.4 +/-9.7 vs 34.8 +/- 9.6 (p<0.001)
    • Time to initial NA (hours) 2.7 +/- 2.1 vs 3.8 +/- 2.9 (p<0.002)
    • Time to abx (hours) 1.4 +/-1.2 vs 2.2 +/-1.8 (p0.001)
    • Volume of IV fluids in 6 hours (litre) 3.4 +/-0.9 vs 3.0 +/-0.9 (p0.003)
    • Volume of IV fluids in 24 hours (L) 6.5 +/-0.8 vs 6.9 +/-0.5 (p<0.001)
    • NA duration (days) 2.4 +/-0.6 vs 3.4 +/- 0.9 (p<0.001)
  • Time to NA initiation corresponds to 28 day mortality (OR)
    • 5% if <1 hour
    • 4% if 1-2 hours
    • 2% if >6 hours
  • OR of death was 1.2 per hour delay
  • Arbitrary 2 hours cut-off drawn by authors and patients split into early and late
  • Early NA group has
    • Less IV fluids in 24 hours 6.2L +/- 0.6 vs 6.9 +/- 0.7 (p<0.001)
    • Less total NA use over 24 hours 29.4mg +/- 9.7 vs 32.8 +/- 10 (p0.013)
    • Shorter duration of hypotension 4.6 hours +/- 1.2 vs 6.1 +/- 1 (p<0.001)
    • Shorter duration of NA use 2.6 days +/- 0.6 vs 2.9 +/- 1. (p<0.001)
    • Higher MAP p<0.05
  • Time to initial NA use, time to abx use, serum lactate, APACHE II score, IV therapy within 6 hours all independent determinants of 28 day mortality
  • Early NA spares fluid volume
  • Results consistent with previous studies RE: time to abx use, APACHE II score and timing of NA use



  • Defined enrolment into study was onset of septic shock in all patients who did not have any other obvious cause of hypotension
  • Representative of the population at QMC (surgical/trauma ITU)
  • Not too specialised, i.e. neuro ITU
  • May not apply to patients in smaller ITU from DGH
  • Medical patients – different pathophysiology (no surgical puncture)


Critical appraisal

  • No mention of morbidity
  • What if there is an element of hypotension related to sedation – does this affect MAP target/timing of NA?
  • What if the patient is on the ward? Patients with persisting hypotension >1 hour prior to admission excluded – only relevant to patients becoming septic in ITU?
  • What if someone was initially fluid responsive then had a septic shower and required NA after 6 hours?
  • No mention of cause of mortality – was it related to hypotensive episode?
  • Authors correctly point out that late use of NA could reflect poor evaluation of disease severity +/- generally poorer global management
  • Study agrees with previous studies RE: time of NA use
  • Retrospective study – unable to identify cause of delay if not documented
  • Relatively small sample size of 213 only
  • Suggest larger study

1 comment so far

  1. quarteredonion on

    Some further points to consider…

    No mention of “sepsis six” per se in this manuscript.

    Study limited to oriental population – see below.

    Reporting of mortality Vs time to noradrenaline is different – see figure 3.

    Findings of this paper seem to contradict the findings from two recent studies;

    Waetcher et al, Critical Care Medicine, 2104,
    Detrimental to start noradrenaline in the first hour without initiating and continuing aggressive fluid loading.


    Anatasit et al, Critical Care Medicine, 2014
    Serious adverse events associated with noradrenaline (and vasopressin) are associated with mortality outcome, and are probably genetically linked.


    There is likely to be confounding from disease severity, early identification, and timely “bundle” treatment”, i.e. those with very bad shock died before catecholamine infusion, those who got catecholamine infusion made it to critical care for lines, fluids, antibiotics, skilled physician, skilled nurses etc.

    Interesting, none the less, but as ever, balance to be struck between intervention ad absurdium and sensible pragmatic administration of fluids, catecholamines, and early identification of therapeutic invitation set points (i.e. threshold of fluid bolus prior to catecholamines, and target blood pressure, and organ performance vs global oxygen delivery).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: