Phenotypically matched vs. traditional screen method for preparing red blood cell units in patients with abnormal placentation: a decision analysis approach.

Abstract

Patients with abnormal placentation are hospitalized approximately 21days prior to scheduled delivery due to risk of peri-partum bleeding. It is required to have at least one red blood cell (RBC) unit ready throughout hospitalization. This can be accomplished by performing a type and cross (T&C) every 72h (traditional method) or providing phenotypically matched RBC units (alternative method). A Markov-based model was created to compare the cost-benefits between these two methods. A patient either undergoes the traditional or alternative method. If the patient already made an antibody and/or has rare phenotypes requiring frozen RBC unit at admission, then she can only enter the traditional method. All patients receive a T&C at the end of the 21st day in order to prepare RBC units for scheduled delivery. Model parameters are derived from literature and institutional experiences. Simulation was run for modelling of the 21-day hospitalization of each patient. The expected cost per RBC unit of the traditional method is significantly higher than the alternative method (876·59 vs. 608·32USD, P-value <0·01). For each RBC unit, 4·85h of technician working time is saved by the alternative method. However, to be cost-beneficial, the alternative method should be used only if the total cost of a RBC unit is <857·67USD. Our model demonstrated that preparing phenotypically matched RBC units at admission is more cost-beneficial comparing to the traditional method of T&C every 72h if the cost for the matched unit is <857·67USD.