Use of an anti-D-alloimmunization kinetics model to correct the interval censored D-alloimmunization rate following red blood cell transfusions

Introduction: The rate of D-alloimmunization amongst RhD-negative recipients of RhD-positive red blood cell (RBC) transfusions is not certain. Recipients with a short duration between the index RhD-positive transfusion and the last antibody detection test that did not show anti-D might become D-alloimmunized in the future. A regression model was developed to predict how often such patients might develop D-alloimmunization in the future to help account for the immunohematological uncertainty that accompanies having short serological follow up periods. Methods: Using the published literature on recipients who were intentionally transfused with RhD-positive RBCs and serially followed with antibody screens, as well as unpublished datasets, a regression model was constructed to demonstrate the timing of D-alloimmunization for recipients who became D-alloimmunized within 6 months following the index transfusion. The model was then applied to a series of RhD-negative hospitalized recipients of at least one unit of RhD-positive RBCs who did not become D-alloimmunized but who had fewer than 6 months of serological follow up to weight their contribution to the D-alloimmunization rate. Results: Overall, the rate of D-alloimmunization was 21/105 (20.0%). There were 39 patients whose last documented antibody screen was performed between 14 days and 6 months after the index RhD-positive transfusion, and these patients were entered into the weighted model. After applying the model, the D-alloimmunization rate rose to 26.3%. Conclusion: Using a weighted model can help reduce the immunohematological uncertainty that accompanies the inclusion of patients with relatively short serological follow up in studies of RBC alloimmunization.