TL;DR
This nationwide study in South Korea investigated blood transfusion patterns in severe COVID-19 patients, finding that transfusions were associated with increased 90-day mortality. The research identifies specific risk factors for transfusion and highlights the need for optimized blood resource management in critical care settings.
What Problem Does This Paper Address?
The COVID-19 pandemic significantly disrupted blood supply chains and increased demand for blood products, particularly for critically ill patients. While it's known that severe COVID-19 can lead to complications requiring transfusions, detailed insights into who receives them, what products are used, and the impact of these transfusions on patient outcomes, especially mortality, were lacking. This paper aims to fill that gap by analyzing real-world data to understand these patterns and identify associated risk factors in a large, representative population.
Methods
The study employed a nationwide, population-based cohort design using claims data from South Korea's National Health Insurance Service-National Health Information Database (NHIS-NHID). They analyzed data from January 2021 to December 2021, focusing on adult patients (≥19 years) diagnosed with severe COVID-19 requiring oxygen therapy. To account for potential biases, propensity score matching (1:2) was used to compare patients who received transfusions with those who did not, adjusting for baseline covariates like age and sex. Statistical analyses included multivariate logistic regression to identify risk factors for transfusion and multivariable Cox regression to assess predictors of 90-day mortality.
Datasets
The primary dataset used was the National Health Insurance Service-National Health Information Database (NHIS-NHID) from the Republic of Korea, covering nearly all residents (approximately 52 million). This dataset provided comprehensive claims data for patients diagnosed with COVID-19.
Key Insights and Ideas
- Transfusion as a Marker of Severity and Risk: The study strongly suggests that blood transfusion in severe COVID-19 patients isn't just a treatment but also a marker of underlying disease severity and a predictor of worse outcomes, specifically increased mortality.
- ECMO as a Major Driver for Transfusion: Patients requiring Extracorporeal Membrane Oxygenation (ECMO) were significantly more likely to receive blood transfusions, highlighting the extreme critical care needs of this subgroup.
- Anticoagulation's Dual Role: While anticoagulants were associated with a reduced risk of transfusion and mortality, their use also implies a higher risk of bleeding, which is a primary indication for transfusion. This points to a complex interplay that warrants further investigation.
- Unexpected Protective Factors: The identification of cerebrovascular diseases, a high Charlson Comorbidity Index (CCI) score, and ICU admission as protective factors against transfusion was surprising and suggests that factors beyond direct disease severity might be at play (e.g., physician discretion, patient management strategies in ICU).
Key Results
- Overall Transfusion Rate: 7.11% of severe COVID-19 patients requiring oxygen therapy received at least one blood product within 3 months of diagnosis.
- Blood Product Usage: Packed red blood cells (pRBCs) were the most frequently transfused product (14,477 units), followed by platelet concentrates (20,532 units), fresh frozen plasma (FFP) (4501 units), and cryoprecipitate (1875 units).
- Mortality Association: Blood transfusion was independently associated with a significant increase in 90-day mortality (Hazard Ratio: 1.36, 95% CI: 1.21–1.53).
- Risk Factors for Transfusion: Significant risk factors included chronic kidney disease, surgery, thrombolysis/thrombectomy, renal replacement therapy, mechanical ventilation, and ECMO.
- Protective Factors for Transfusion: Conversely, cerebrovascular diseases, a CCI score of ≥6, ICU admission, and anticoagulant use were identified as protective factors against transfusion.
Limitations and Future Work
The authors acknowledge several limitations:
- Clinician Discretion: The decision to transfuse was based on physician discretion, leading to potential variations in practice and an inability to derive an optimized transfusion algorithm.
- Data Source Limitations: Reliance on NHIS claims data meant a lack of detailed clinical information, such as specific transfusion indications, bleeding status, or precise dosing regimens. This could lead to misclassification of disease severity or transfusion context.
- Confounding Factors: The potential inclusion of patients with "do-not-resuscitate" orders who weren't actively transfused, despite indications, could introduce confounding.
Future work should focus on:
- Validating Transfusion Strategies: Establishing and validating transfusion strategies to improve treatment outcomes in critically ill COVID-19 patients.
- Comparative Analysis: Investigating whether severe COVID-19 patients require more transfusions than those with other critical illnesses.
- Long-Term Effects: Further research into the long-term effects of transfusions on patient outcomes.
- Interplay of Anticoagulation and Bleeding: Exploring the complex relationship between anticoagulation, bleeding events, and blood product use.
Summary updated