No Benefit From Platelet Transfusion for Gastrointestinal Bleeding in Patients Taking Antiplatelet Agents

Liam Zakko, Tarun Rustagi, Maura Douglas, Loren LaineClinical 2016 · DOI

TL;DR

This study investigated whether giving platelet transfusions to patients with gastrointestinal bleeding (GIB) who are already taking antiplatelet medications (like aspirin or clopidogrel) helps prevent rebleeding or improves outcomes. The findings suggest that platelet transfusions do not reduce rebleeding and are actually associated with a significantly higher risk of death, leading the authors to recommend against their use in this patient population.

What Problem Does This Paper Address?

Patients taking antiplatelet agents are at increased risk for gastrointestinal bleeding (GIB). Current guidelines and expert opinion often suggest platelet transfusions for serious GIB in these patients, even if their platelet counts are normal. The rationale is to provide functional platelets to stop bleeding and prevent rebleeding. However, there's a significant concern that these transfusions might paradoxically increase the risk of cardiovascular events (like heart attack or stroke) due to the sudden influx of functional platelets. This paper aims to clarify whether the potential benefits of platelet transfusions in preventing GIB rebleeding outweigh the potential risks, particularly cardiovascular events and mortality.

Methods

The researchers conducted a retrospective cohort study using data from Yale-New Haven Hospital between 2008 and 2013. They identified patients who had GIB, were taking antiplatelet agents, and had a platelet count above 100 x 10^9/L.

  • Cases: Patients who received platelet transfusions (n=204).
  • Controls: Patients who did not receive platelet transfusions (n=204).

Controls were matched to cases based on sex, age, and GIB location. The primary outcome measured was recurrent GIB during the hospital admission. Secondary outcomes included major adverse cardiovascular events (MACE), mortality, and length of hospital stay. Multivariable regression analyses were used to adjust for differences in baseline characteristics between the two groups.

Datasets

The study utilized a retrospective cohort of patients admitted to Yale-New Haven Hospital between 2008 and 2013. No new datasets were created or introduced.

Key Insights and Ideas

  • Platelet transfusions don't help prevent rebleeding: Despite the theoretical benefit, the study found no evidence that platelet transfusions reduced the risk of recurrent GIB in patients already on antiplatelet therapy.
  • Increased mortality is a significant concern: The most striking finding was a substantially higher mortality rate in patients who received platelet transfusions compared to those who did not, even after accounting for various confounding factors.
  • GIB-related deaths are a major driver of mortality: A notable portion of the increased deaths in the transfusion group were attributed to GIB itself, suggesting that transfusions might not only fail to stop bleeding but could potentially worsen it or its consequences.
  • Observational bias is acknowledged but unlikely to explain the full effect: While the authors recognize that observational studies can have residual bias, the magnitude of the observed difference in mortality suggests a potential causal link.
  • Current guidelines may lack strong evidence: The paper points out that existing guidelines recommending platelet transfusions for GIB in this context may not be well-supported by robust evidence.

Key Results

  • Recurrent GIB: The adjusted odds ratio for recurrent GIB was 1.47 (95% CI, 0.73-3.05), indicating no statistically significant difference between cases and controls.
  • Mortality: The adjusted odds ratio for mortality was 5.57 (95% CI, 1.52-27.1), showing a significant increase in the risk of death for patients receiving platelet transfusions.
  • Major Adverse Cardiovascular Events (MACE): While univariate analysis showed a higher rate of MACE in the transfusion group, the multivariable analysis did not show a significant difference (adjusted OR, 1.35; 95% CI, 0.74-2.49).
  • Length of Stay: The adjusted odds ratio for a hospital stay longer than 4 days was 1.13 (95% CI, 0.71-1.79), indicating no significant difference.
  • Red Cell and FFP Transfusions: Patients receiving platelet transfusions also received significantly more red blood cell and fresh frozen plasma transfusions, suggesting they presented with more severe bleeding.

Limitations and Future Work

  • Retrospective Nature: The study's retrospective design is a primary limitation. While efforts were made to control for confounding factors, statistical adjustments may not fully account for all differences between the groups, and residual bias is possible.
  • Generalizability: The study population was heterogeneous in terms of GIB sources. The results may not be generalizable to patients with variceal bleeding, as this was a small proportion of the study cohort. Additionally, the findings might differ for outpatients versus inpatients.
  • Unmeasured Confounders: Despite extensive data collection, there might be unmeasured factors that influenced physician decisions regarding transfusions or patient outcomes.

The authors suggest that further prospective studies are needed to confirm these findings and better assess the role of platelet transfusions in patients with GIB who are taking antiplatelet agents. They emphasize that practices based on conjecture should be avoided if they do not demonstrate clear benefit and may even cause harm.

Summary updated

BibTeX
@inproceedings{zakko2016no,
  author = {Liam Zakko and Tarun Rustagi and Maura Douglas and Loren Laine},
  title = {{No Benefit From Platelet Transfusion for Gastrointestinal Bleeding in Patients Taking Antiplatelet Agents}},
  booktitle = {Clinical},
  year = {2016},
  doi = {10.1016/j.cgh.2016.07.017},
}
Figure