A Single-Center Description of Pediatric Transfusion Reactions and Preventable Patient Harm

This study was approved by the institutional review board of our institution. Data were obtained retrospectively for pediatric patients who had a documented transfusion reaction at our institution from January 2018 through December 2019.

Pediatric patients aged <21 years who had a documented acute transfusion reaction (ATR) to blood products (red blood cells [RBCs], platelets, fresh-frozen plasma [FFP], and cryoprecipitate) were included in this study. Delayed transfusion reactions and reactions to other cellular products, including hematopoietic progenitor cells, or plasma derivatives were not included.

Electronic medical records, including the transfusion reaction reports, were reviewed. Demographic data (age and sex), current diagnosis, and transfusion trigger were collected. The transfusion reaction reports included description, definition, severity, and imputability of the reaction according to NHSN guidelines.⁴  Transfusion indication was retrospectively categorized as indicated, not indicated, or questionably indicated on the basis of chart review of reaction reports and clinical context. Transfusions defined as indicated clearly met institution and AABB transfusion guidelines.^12,13,16,17  Transfusions defined as not indicated clearly did not meet these guidelines.

Transfusions that were not indicated but had documented clinical rationale were independently reviewed by 2 transfusion medicine attending physicians. If the documented rationale was not reasonable, reactions were classified as not indicated. However, if there was reasonable rationale that would have benefited from a real-time discussion with transfusion medicine, the transfusion was classified as questionably indicated. For example, RBC transfusion would be classified as not indicated in a patient with a hemoglobin (Hb) above transfusion threshold and documented rationale to “increase oxygen capacity.” However, RBC transfusion would be questionably indicated if the team transfused for a threshold Hb of 8 mg/dL instead of 7 mg/dL and documented that the patient had a history of symptomatic anemia <8 mg/dL or that the patient recently received myelosuppressive chemotherapy and was expected to become increasingly anemic.

Patient and transfusion reaction characteristics were summarized by using descriptive statistics with Microsoft Excel (Microsoft 365 Version 16.39 [2020]; Microsoft Corporation, retrieved from https://office.microsoft.com/excel).

Transfusion reaction percentage was calculated as number of transfusion reactions per total number of transfusions for the same population and time period. Rate of transfusion reactions was calculated as number of transfusion reactions per 100 000 transfusions. PPH was calculated as a percentage (number of transfusion reactions due to inappropriate blood transfusions per total number of transfusion reactions) and as a rate (number of transfusion reactions due to inappropriate blood transfusions per 100 000 transfusions).

A total of 155 ATRs occurred in 106 patients over the 2-year period of the study. In Table 1, we summarize patient characteristics.

The average rate of transfusion reaction was 544 of 100 000 transfusions (561 in 2018 and 528 in 2019). The majority of reactions resulted from transfusions that were indicated (65%, 100 of 155). Thirty-six reactions resulted from questionably indicated transfusions (23%, 36 of 155). Twenty reactions resulted from not-indicated transfusions (12%, 19 of 155) and represent PPH with a reaction rate 67 of 100 000 transfusions. Ninety-five percent (18 of 19) of not-indicated reactions were classified as nonsevere, indicating mild patient harm. The majority of these were febrile nonhemolytic transfusion reactions (FNHTRs) (47%, 9 of 19) and mild ATRs (32%, 6 of 19). The rates and characteristics of transfusion reactions are summarized in Table 2.

The most common types of reactions were FNHTRs (60%, 93 of 155) and ATRs (26%, 41 of 155). Table 2 reveals reaction rates for the study period and products associated with each reaction. Pretransfusion triggers were Hb 7.91 ± 2.07 g/dL (range 3.9–13.6 g/dL) for RBC transfusions, platelet count 25.78 ± 41.11 × 10³/μL (range 1–334 × 10³/μL) for platelet transfusions, and international normalized ratio 2.08 ± 0.28 (range 1.6–2.2) for FFP transfusions. Two multiproduct transfusions were performed for triggers of Hb 6.5 g/dL and platelets 29 × 10³/μL and Hb 6.1 g/dL and INR 2.7.

NHSN definition (diagnostic) criteria, severity, and imputability are illustrated in Table 3. The majority (83.2%) of transfusions were definitive by NHSN diagnostic criteria. Ninety-three percent of reactions were classified as nonsevere. Imputability is a term used by the NHSN to describe the degree to which a reaction is attributable to its associated transfusion. Thirty-nine percent of reactions had definite imputability, 19% were probable, 36% were possible, and a minority of reactions were of doubtful or indeterminate imputability.

We reviewed all pediatric transfusions resulting in reactions over a 2-year period at our institution and applied our institutional guidelines, pediatric hematology-oncology Choosing Wisely Guidelines, and AABB guidelines to categorize reactions as indicated, not indicated, or questionably indicated. On the basis of these peer-reviewed institution guidelines, the majority of transfusions were indicated, 23% were questionably indicated, and 12% were not indicated. Reactions resulting from indicated transfusions are not preventable, whereas reactions resulting from not-indicated transfusions represent PPH that could be prevented by improved transfusion stewardship. The rate of PPH attributed to inappropriate transfusions was 67 of 100 000 transfusions. To our knowledge, this is the first study relating appropriateness of pediatric transfusions to PPH.

In addition, this study adds to the body of literature on pediatric transfusion reactions by describing the characteristics of patients and transfusion reactions. The overall reaction rate of 544 of 100 000 transfusions is comparable to previously published rates of 620 and 833 of 100 000 transfusions^5,6  in pediatric populations. This adds evidence that reaction rates are approximately twice as great in the pediatric population compared with the adult population.^2,5,6 

One of the limitations of this study is its retrospective nature. A real-time audit may have resulted in further clarification for transfusions with questionable indication. For this reason, only the reactions that were clearly not indicated (and not those of questionable indication) were used to assess PPH. PPH was possibly underestimated as a result. If questionably indicated transfusions had been included in our definition of PPH, PPH would have been 35% rather than 12%. Additional prospective and real-time studies may be used to clarify this limitation. Complicating this is lack of universally standardized pediatric transfusion guidelines. However, each institution is required to define transfusion guidelines on the basis of the most-updated literature, review compliance, and measure of patient impact. An additional limitation was that delayed transfusion reactions were not included in this study.

Despite limitations, quantifying PPH from the perspective of transfusion indication is potentially a novel quality indicator of transfusion practice. The importance of developing adequate transfusion guidelines and documentation cannot be underestimated. Standardization of reaction reporting is critical for interinstitutional comparison and furthers opportunities for practice improvement.

On the basis of our experience, 12% to 35% of pediatric transfusion reactions at our institution were potentially preventable by improved transfusion practices. Further improvements and research in pediatric transfusion medicine are needed to facilitate this. Measures to reduce inappropriate transfusions and PPH in pediatric populations include standardization of evidence-based transfusion thresholds for pediatric patients, increased education and feedback by the transfusion medicine service to ordering providers regarding transfusion appropriateness and potential harms, and real-time audits of pediatric transfusion indications.

Inappropriate transfusions that result in reactions in the pediatric population represent PPH that is potentially actionable via physician education, real-time audits of transfusion indications, and standardized pediatric transfusion thresholds. Future efforts should be made to standardize and better educate pediatric providers on transfusion guidelines to ensure patient safety.