Effect of post-storage filters vs. pre-storage filters for leukoreduction of blood components on clinical outcomes: a systematic review and meta-analysis.

Background: Leukoreduction has been used to limit the risk of adverse events. The most commonly used methodology is filtration (pre- or post-storage). However, whether pre-storage filtration is better than post-storage filtration needs to be clearly defined, particularly for countries that still use post-storage filtration. This study aimed to synthesize the best available evidence on the effectiveness of pre-storage filters compared with post-storage filters for transfusion reactions, for the occurrence of infections, for the length of hospital stay, and for the death of patients undergoing leukoreduced transfusion.

Methods: We searched the MEDLINE (PubMed), CINAHL (EBSCO), PsycINFO (APA), Scopus (Elsevier), The Cochrane Library (J. Wiley), Web of Science Core Collection (Clarivate Analytics), Embase (Elsevier), and LILACS (VHL) databases and gray literature for eligible studies in August 2020 and updated the search in October 2023. The Joanna Briggs Institute critical assessment tools were applied to analyze the quality appraisal of the studies. GRADE was used to determine the certainty of the evidence.

Results: The meta-analysis showed that pre-storage filtration was a protective factor for the occurrence of febrile non-hemolytic transfusion reaction in red blood cells (RR 0.49, 95% CI 0.41-0.59) and platelet concentrate transfusions (RR 0.16, 95% CI 0.12-0.22). The same did not occur for post-surgical infection after platelet concentrate transfusions (RR 0.82, 95% CI 0.65-1.04). Only one study analyzed the length of hospital stay and showed no significant difference between patients who received leukoreduced transfusions according to the type of filter used. According to the GRADE criteria, the certainty of the evidence for febrile non-hemolytic transfusion reactions was low for red blood cells and very low for platelet concentrate due to the high risk of bias. Infection was a low risk due to imprecision.

Conclusions: The results of this review showed that the certainty of recommending the best type of filter (pre- or post-storage) for the benefit of the outcomes analyzed is still fragile; therefore, more robust evidence is needed.

Systematic Review Registration: PROSPERO CRD42020192202.