TRISS: Lower versus Higher Hgb Threshold for Transfusion in Septic Shock

<— Back to Education Page

Holst L et al. "Lower versus Higher Hemoglobin Threshold for Transfusion in Septic Shock". The New England Journal of Medicine. 2014; 371:1381-1391 .

Link to Article:

Clinical Question

In patients with septic shock, does blood transfusion at a lower or higher hemoglobin threshold improve mortality?

  • Type of Trial: RCT

  • Setting: 32 general ICUs in Scandinavia from Dec 2011 - Dec 2013


  • Population:

    • All patients with septic shock and Hgb<9

      • Excluded: patients with ACS, previous adverse reaction to transfusion, life threatening bleeding, acute burn, withdrew from active therapy

    • In comparing intervention/control group no significant difference in baseline characteristics; most common source of sepsis lungs (53.2% vs 52.2%)

  • Intervention:

    • Transfusion threshold of <7 g/dl

  • Control:

    • Transfusion threshold of <9 g/dl

    • In both intervention/control:

      • Pts given single unit crossmatched, prestorage leukoreduced red cells when met threshold

      • Intervention period was entire ICU stage (max of 90d after randomization)

      • If life threatening bleeding, ischemia or ECMO required then attending doctor could transfuse Hgb at their own threshold

  • Outcome:

    • Primary outcome: death by 90 days

      • No significant difference (43.0% intervention vs 45.0% control, P=0.44)

    • Secondary outcomes:

      • Median number of blood transfusions - lower in lower threshold (43.0% intervention vs 45.0% control, P=0.44)

      • Number of patients that did not undergo transfusion - significantly more in lower threshold (43.0% intervention vs 45.0% control, P=0.44)

      • No significant difference in:

        • use of life support at days 5/ 14/ 28 (64.4%/36.8%/16.1% intervention 62.2%/36.8%/19.9% control, P=0.47/0.95/0.14)

        • alive wo vasopressor or inotropic support (73.0% intervention vs 75.0% control, P=0.93)

        • alive w/o mechanical ventilation (65.0% intervention vs 67.0% control, P=0.49)

        • alive w/o renal replacement therapy (85.0% intervention vs 83.0% control, P=0.54)

        • percentage of days alive and out of hospital (30d intervention vs 31d control, P=0.89)

        • Severe complications: no significant difference

          • Ischemic event: 7.2% intervention vs 8.0% control, P=0.64

          • Severe adverse rxn: 0% intervention vs 0.2% control, P=1.00

    • Sub-group analysis: no sig difference in primary outcome based on presence/abscence of chronic CV disease, age> or < 70, SAPS II of > or <53


In patients with septic shock, mortality and rates of ischemic events were similar in those assigned to a blood transfusion at a higher vs. lower threshold .


  • Randomized

  • Blinding of investigators assessing primary outcome

  • Multi-center


  • Non-blinding of staff (although unrealistic to achieve)

  • Protocol violations: patients were more likely to be given transfusion as either a suspension or violation if target <7 and more likely to have transfusion if threshold <9

Take Note

  • Patients with ACS were excluded; further research is needed in this group

  • Having a lower transfusion threshold does NOT result in difference in mortality or ischemic events; but it does remove risk of infectious, immune complications and conserves limited resource


Rachel MulderComment