Target Hemoglobin in Sepsis
In patients with sepsis, transfuse red blood cells only when hemoglobin falls below 7.0 g/dL and target a hemoglobin concentration of 7.0-9.0 g/dL once tissue hypoperfusion has resolved. 1
Restrictive Transfusion Strategy
The Surviving Sepsis Campaign guidelines consistently recommend a restrictive transfusion approach across multiple iterations:
- Transfusion threshold: Hemoglobin <7.0 g/dL 1
- Target range: 7.0-9.0 g/dL 1
- This recommendation carries a Grade 1B (strong recommendation, moderate quality evidence) 1
The evidence supporting this restrictive strategy comes from the landmark TRISS trial, which randomized 998 patients with septic shock to either a lower threshold (transfuse at Hb ≤7 g/dL) or higher threshold (transfuse at Hb ≤9 g/dL). 2 The study found no difference in 90-day mortality (43.0% vs 45.0%, P=0.44), ischemic events, or need for life support between groups, while the lower threshold group received significantly fewer transfusions (median 1 unit vs 4 units). 2
Important Exceptions Requiring Higher Thresholds
Do not apply the 7.0 g/dL threshold in the following circumstances: 1
- Active myocardial ischemia or acute coronary syndrome - consider higher thresholds 1
- Severe hypoxemia - may require higher hemoglobin for oxygen delivery 1
- Acute hemorrhage - transfuse based on ongoing blood loss 1
- Documented ischemic coronary artery disease - individualize threshold 1
During the initial resuscitation phase (first 6 hours) when tissue hypoperfusion persists and ScvO2 remains low, early goal-directed therapy protocols have historically used higher targets (hematocrit 30%, approximately Hb 10 g/dL), though this contrasts with the restrictive strategy applied after stabilization. 1
Clinical Context and Nuances
Timing matters critically: The restrictive 7.0 g/dL threshold applies specifically once tissue hypoperfusion has resolved. 1 During active shock with ongoing hypoperfusion, clinical judgment regarding oxygen delivery needs takes precedence.
Mortality correlates with severity of anemia: While the restrictive strategy is safe, observational data shows that lower initial hemoglobin levels (<9.0 g/dL) at presentation correlate with increased mortality in septic shock patients, with odds ratios progressively increasing as hemoglobin drops below 9.0 g/dL (OR 2.35 for Hb <7.0 g/dL). 3 However, this association does not mean transfusion to higher targets improves outcomes—the TRISS trial definitively showed it does not. 2
Fluid balance interactions: In patients with moderate anemia (Hb 7-10 g/dL), positive fluid balance may be associated with increased mortality, whereas in non-anemic patients, fluid resuscitation shows better tolerance. 4 This suggests careful attention to both hemoglobin levels and volume status during resuscitation.
What NOT to Do
- Do not use erythropoietin for sepsis-associated anemia—it provides no mortality benefit and does not improve clinical outcomes (Grade 1B). 1
- Do not transfuse based solely on laboratory values in stable patients without bleeding or the extenuating circumstances listed above 1
- Avoid liberal transfusion strategies (targeting Hb >9-10 g/dL) as they expose patients to unnecessary transfusion risks without mortality benefit 2
Common Pitfalls
Pitfall #1: Transfusing to "normal" hemoglobin levels (>10 g/dL) in stable septic patients. The TRISS trial showed this doubles transfusion requirements without improving survival. 2
Pitfall #2: Applying the restrictive threshold during active resuscitation with unresolved tissue hypoperfusion. The 7.0 g/dL threshold is for the post-resuscitation phase. 1
Pitfall #3: Ignoring cardiac comorbidities. Patients with acute coronary syndromes or severe coronary disease may require higher thresholds, though specific targets for these populations remain understudied. 1, 5
Pitfall #4: Using fresh frozen plasma to correct coagulation abnormalities in non-bleeding patients—this is not recommended (Grade 2D). 1