VILLANUEVA: Restrictive vs Liberal Transfusion in Upper GI Bleeding (2013)

“As compared with a liberal transfusion strategy, a restrictive strategy significantly improved outcomes in patients with acute upper gastrointestinal bleeding.”

— The Villanueva et al. Study Group

1. Publication Details

  • Trial Title: Transfusion Strategies for Acute Upper Gastrointestinal Bleeding.
  • Citation: Villanueva C, Colomo A, Bosch A, et al. Transfusion strategies for acute upper gastrointestinal bleeding. N Engl J Med. 2013;368(1):11-21. doi:10.1056/NEJMoa1211801.
  • Published: January 3, 2013, in The New England Journal of Medicine.
  • Author: Càndid Villanueva, M.D.
  • Funding: Spanish Ministry of Science and Innovation and others.

2. Keywords

Gastrointestinal Hemorrhage, Blood Transfusion, Anemia, Hemoglobin, Transfusion Trigger, Liver Cirrhosis, Portal Hypertension.

3. The Clinical Question

In adult patients with acute upper gastrointestinal bleeding (Population), does a restrictive red blood cell transfusion strategy (hemoglobin trigger of 7 g/dL) (Intervention) compared to a liberal transfusion strategy (hemoglobin trigger of 9 g/dL) (Comparison) reduce 45-day all-cause mortality (Outcome)?

4. Background and Rationale

  • Existing Knowledge: The optimal transfusion strategy for patients with acute upper GI bleeding was unknown. While transfusion can restore hemodynamic stability, it can also increase portal pressure in patients with cirrhosis, potentially worsening or re-starting variceal bleeding. The TRICC trial had shown the safety of a restrictive strategy in general ICU patients, but a dedicated trial in this specific high-risk population was needed.
  • Knowledge Gap: It was unclear if a restrictive or liberal transfusion strategy was superior in patients with acute upper GI bleeding, particularly in the large subgroup of patients with liver cirrhosis.
  • Proposed Hypothesis: The authors hypothesized that a restrictive transfusion strategy would be superior to a liberal strategy in improving clinical outcomes in patients with acute upper GI bleeding.

5. Study Design and Methods

  • Design: A prospective, single-center, randomized, controlled trial.
  • Setting: A single tertiary care university hospital in Spain.
  • Trial Period: Enrollment from June 2008 to December 2011.
  • Population:
    • Inclusion Criteria: Adult patients (≥18 years) presenting with severe acute upper gastrointestinal bleeding (hematemesis or melena).
    • Exclusion Criteria: Massive exsanguinating bleeding, lower GI bleeding, recent trauma or surgery, or acute coronary syndrome.
  • Intervention: A restrictive transfusion strategy. RBCs were transfused only when the hemoglobin level dropped below 7 g/dL, with a target of maintaining it between 7 and 9 g/dL.
  • Control: A liberal transfusion strategy. RBCs were transfused when the hemoglobin level dropped below 9 g/dL, with a target of maintaining it between 9 and 11 g/dL.
  • Management Common to Both Groups: All patients received standard medical and endoscopic therapy for upper GI bleeding.
  • Power and Sample Size: The trial was powered to detect a 10% absolute difference in the incidence of further bleeding, requiring 900 patients.
  • Outcomes:
    • Primary Outcome: All-cause mortality at 45 days.
    • Secondary Outcomes: Included rates of further bleeding, need for rescue therapy, and adverse events.

6. Key Results

  • Enrollment and Baseline: 921 patients were randomized (461 to the restrictive group, 460 to the liberal group). The groups were well-matched. Approximately 93% of patients had liver cirrhosis.
  • Trial Status: The trial was completed as planned.
  • Primary Outcome: All-cause mortality at 45 days was significantly lower in the restrictive-strategy group than in the liberal-strategy group (5% vs 9%; P=0.02).
  • Secondary Outcomes: The incidence of further bleeding was significantly lower in the restrictive group (10% vs 16%; P=0.01). The need for rescue therapy (balloon tamponade or TIPS) was also lower.
  • Adverse Events: The rate of transfusion reactions was lower in the restrictive group. There was a trend toward more adverse events, particularly circulatory overload, in the liberal group.

7. Medical Statistics

  • Analysis Principle: An intention-to-treat analysis was performed.
  • Statistical Tests Used: The primary outcome was analyzed using the log-rank test.
  • Primary Outcome Analysis: Time to death from any cause up to 45 days was compared between the two groups using a Kaplan-Meier analysis.
  • Key Statistic(s) Reported: Hazard Ratio (HR) for death at 45 days with the restrictive strategy: 0.55 (95% CI, 0.33 to 0.92; P=0.02).
  • Interpretation of Key Statistic(s):
    • Hazard Ratio (HR):
      • Formula: Conceptually, HR represents the instantaneous risk of death in the intervention group relative to the control group.
      • Calculation: The paper reports the HR as 0.55.
      • Clinical Meaning: An HR of 0.55 means there was a 45% lower hazard of death at any given time in the restrictive transfusion group compared to the liberal group.
    • Confidence Interval (CI):
      • Formula: Conceptually, CI = (Point Estimate) ± (Margin of Error).
      • Calculation: The reported 95% CI was 0.33 to 0.92.
      • Clinical Meaning: Since the 95% CI ranges from a 67% benefit to an 8% benefit and does not cross 1.0, the result is statistically significant.
    • P-value:
      • Calculation: The reported p-value was 0.02.
      • Clinical Meaning: The p-value of 0.02 is less than the conventional threshold of 0.05, indicating that the observed difference is unlikely to be due to chance.
  • Clinical Impact Measures:
    • Absolute Risk Reduction (ARR):
      • Formula: ARR = (Risk in Control Group) – (Risk in Intervention Group).
      • Calculation: ARR = 9.1% – 5.0% = 4.1%.
      • Clinical Meaning: For every 100 patients treated with a restrictive strategy, about 4 fewer died by day 45.
    • Number Needed to Treat (NNT):
      • Formula: NNT = 1 / ARR.
      • Calculation: NNT = 1 / 0.041 = 24.4.
      • Clinical Meaning: Approximately 24 patients with acute upper GI bleeding need to be treated with a restrictive transfusion strategy to prevent one additional death at 45 days.
  • Subgroup Analyses: The benefit of the restrictive strategy was most pronounced in the large subgroup of patients with cirrhosis and Child-Pugh class A or B disease.

8. Strengths of the Study

  • Study Design and Conduct: This was a large, randomized controlled trial that provided high-quality evidence on a critical clinical question.
  • Patient-Centered Outcomes: The primary outcome of 45-day mortality is a strong, patient-centered endpoint.
  • Focused Population: The trial focused on a specific, high-risk population where the pathophysiology of transfusion was particularly relevant.

9. Limitations and Weaknesses

  • Internal Validity (Bias): The study was unblinded, which could introduce performance bias.
  • External Validity (Generalizability): The most significant limitation is that this was a single-center study. Additionally, over 90% of the patients had cirrhosis, so the results are most applicable to this group and may not be generalizable to patients with non-variceal bleeding without liver disease.
  • Other: The trial was powered for a secondary outcome (rebleeding), not the primary outcome of mortality.

10. Conclusion of the Authors

“In patients with acute upper gastrointestinal bleeding, a restrictive transfusion strategy significantly improved outcomes as compared with a liberal transfusion strategy.”

11. To Summarize

  • Impact on Current Practice: This was a landmark, practice-changing trial. It provided the first strong evidence that for patients with acute upper GI bleeding (especially those with cirrhosis), a liberal transfusion strategy is not just unnecessary but is actively harmful. It established a restrictive transfusion strategy (trigger of 7 g/dL) as the definitive standard of care for this population.
  • Specific Recommendations:
    • Patient Selection: For adult patients with acute upper GI bleeding, particularly those with known or suspected cirrhosis.
    • Actionable Intervention: Adopt a restrictive transfusion strategy, using a hemoglobin threshold of 7.0 g/dL as the trigger for transfusion.
    • Expected Benefit: A significant reduction in 45-day mortality, with an NNT of approximately 24, as well as a reduction in the risk of further bleeding.
  • What This Trial Does NOT Mean: This trial does not mean the 7.0 g/dL trigger applies to all patients. The results are less certain for patients with non-variceal bleeding without liver disease or those with active, exsanguinating hemorrhage.
  • Implementation Caveats: The findings provide a clear directive to avoid “cosmetic” transfusions to a higher hemoglobin target in this patient population.

12. Context and Related Studies

  • Building on Previous Evidence: This trial provided crucial, specific evidence that extended the findings of the TRICC trial (1999) to the unique population of patients with upper GI bleeding and portal hypertension, where the pathophysiology of transfusion is different.
  • Influence on Subsequent Research: The findings of this trial have been highly influential and were largely confirmed by the subsequent UK multicenter TRIGGER trial (2015). These studies have cemented the restrictive approach in international guidelines.

13. Unresolved Questions & Future Directions

  • Unresolved Questions: What is the optimal transfusion strategy for patients with non-variceal bleeding who do not have cirrhosis? What is the role of transfusion in patients with acute coronary syndrome who present with a GI bleed?
  • Future Directions: Research continues to refine transfusion thresholds for specific patient subgroups and to investigate the role of other hemostatic agents.

14. External Links

15. Framework for Critical Appraisal

  • Clinical Question: The question was of fundamental importance, addressing a common and high-stakes clinical scenario where practice was based on dogma rather than evidence.
  • Methods: The randomized controlled design was appropriate. The single-center nature is the main limitation, but the large sample size and robust findings mitigate this.
  • Results: The trial had a clear and statistically significant positive result for its primary, patient-centered outcome of mortality, as well as for key secondary outcomes.
  • Conclusions and Applicability: The authors’ conclusion is strongly supported by the data. The results are highly applicable, particularly to the large population of patients with cirrhosis and variceal bleeding, and have definitively established the standard of care.

16. Disclaimer and Contact

This summary is provided by the Academic Committee of ESBICM (ACE) to facilitate the understanding of this study; readers are advised to refer to the original trial document for a deeper understanding. If you find any information incorrect, or missing, or it needs an update or have a request for a specific critical care trial summary, kindly write to us at academics[at]esbicm.org.

Scroll to Top