BACKGROUND: Erythrocytapheresis is used to prevent acute chest syndrome and stroke in patients with sickle cell disease (SCD). However, such regimens are associated with significant risks, such as iron overload and potential exposure to transfusion-transmitted infectious diseases. Computer modeling of erythrocytapheresis procedures may help optimize treatments and minimize risks. STUDY DESIGN AND METHODS: Mathematical models based upon material balance equations and patient-specific statistical analyses were developed to estimate HbS levels immediately after erythrocytapheresis and immediately before the next treatment. The equations were incorporated into a software application that was used to model the effects of various treatment values on four patients treated with 90 erythrocytapheresis procedures. RESULTS: Immediate postprocedure HbS values were accurately estimated with correlations between measured and calculated values ranging from R(2) = 0.83 to 0.96. Estimates of HbS just before the next treatment correlated well in three patients (R(2) = 0.71 to 0.83) but poorly in one (R(2) = 0.28 to 0.46). Varying the treatment values by computer simulation led to a wide variation in the number of RBC units and the net RBC volume transfused. CONCLUSION: Computer modeling of erythrocytapheresis can be used to optimize chronic treatment regimens for SCD patients and potentially to minimize the risks of overtransfusion.
BACKGROUND: Erythrocytapheresis is used to prevent acute chest syndrome and stroke in patients with sickle cell disease (SCD). However, such regimens are associated with significant risks, such as iron overload and potential exposure to transfusion-transmitted infectious diseases. Computer modeling of erythrocytapheresis procedures may help optimize treatments and minimize risks. STUDY DESIGN AND METHODS: Mathematical models based upon material balance equations and patient-specific statistical analyses were developed to estimate HbS levels immediately after erythrocytapheresis and immediately before the next treatment. The equations were incorporated into a software application that was used to model the effects of various treatment values on four patients treated with 90 erythrocytapheresis procedures. RESULTS: Immediate postprocedure HbS values were accurately estimated with correlations between measured and calculated values ranging from R(2) = 0.83 to 0.96. Estimates of HbS just before the next treatment correlated well in three patients (R(2) = 0.71 to 0.83) but poorly in one (R(2) = 0.28 to 0.46). Varying the treatment values by computer simulation led to a wide variation in the number of RBC units and the net RBC volume transfused. CONCLUSION: Computer modeling of erythrocytapheresis can be used to optimize chronic treatment regimens for SCDpatients and potentially to minimize the risks of overtransfusion.