Transfusion thresholds and other strategies for guiding red blood cell transfusion.

Summary

Across 61 trials (27,639 participants), restrictive hemoglobin thresholds (7–8 g/dL) reduced red cell exposure by 42% without increasing 30‑day mortality overall. Exceptions emerged: restrictive strategy lowered mortality in gastrointestinal bleeding, while liberal strategy improved unfavorable neurologic outcomes in neurocritical patients. Transfusion reactions were more frequent with liberal strategies; pediatric findings were broadly consistent but less certain.

Key Findings

• Restrictive strategies reduced receipt of ≥1 RBC unit by 42% across 61 trials (RR 0.58, 95% CI 0.52–0.65).
• No overall difference in 30-day mortality between restrictive and liberal thresholds (RR 1.01, 95% CI 0.90–1.14).
• Gastrointestinal bleeding favored restrictive strategy for 30-day mortality (RR 0.63, 95% CI 0.42–0.95).
• Neurocritically ill patients had better 6–12 month neurologic outcomes with liberal strategies (RR 1.14 for unfavorable outcomes).
• Transfusion reactions were less frequent with restrictive approaches (Peto OR 0.47).

Clinical Implications

Adopt restrictive transfusion thresholds (7–8 g/dL) for most adults to reduce exposure without harming mortality, while considering liberal thresholds in neurocritical brain injury and restrictive strategies in GI bleeding. Implement stewardship to reduce transfusion reactions and explore physiologic triggers in selected contexts.

Limitations

• High heterogeneity in transfusion exposure outcomes (I²=97%) reflecting context diversity
• Physiologic trigger trials were heterogeneous precluding meta-analysis; pediatric certainty lower

Future Directions

Define context-specific thresholds (e.g., brain injury, acute MI), integrate physiologic triggers with hemoglobin thresholds, and prioritize patient-centered outcomes (function/quality of life) beyond mortality.