BACKGROUND: Anticoagulant concentration in plasma units is extremely variable. Understanding the underlying causes of this variability could help personalise plasmapheresis procedures in order to optimise the risk-benefit ratio. We studied the association between anticoagulant solution A (usually ACD-A, Citrate Dextrose Solution A) volume in plasma units and donor characteristics to build a model to determine the needed weight of the final plasma unit to have an 80% probability of reaching 600 mL net plasma. MATERIALS AND METHODS: We experimentally measured ACD-A in 296 plasma units from an Italian blood donor centre, where machines are set for the collection of 700 g of plasma. Next, we built a statistical model to predict how the final volume of the unit should be set to obtain 50%, 80% or 90% probability of having at least 600 mL net plasma. RESULTS: ACD-A volume was associated with haemoglobin, total proteins and triglycerides. Donors with low haemoglobin reach an 80% probability of at least 600 mL net plasma with units of approximately 690 g, while 720 g are needed for donors with high haemoglobin levels. For total proteins and triglycerides, plasma units may vary within a range of ±20 g. DISCUSSION: Our model, based on easily measurable individual characteristics, makes it possible to customise plasmapheresis procedures by determining the blood volume to be processed for each donor. Tailored plasma donations might result in both a reduction in adverse events and an increase in the quality of collected plasma.
BACKGROUND: Anticoagulant concentration in plasma units is extremely variable. Understanding the underlying causes of this variability could help personalise plasmapheresis procedures in order to optimise the risk-benefit ratio. We studied the association between anticoagulant solution A (usually ACD-A, Citrate Dextrose Solution A) volume in plasma units and donor characteristics to build a model to determine the needed weight of the final plasma unit to have an 80% probability of reaching 600 mL net plasma. MATERIALS AND METHODS: We experimentally measured ACD-A in 296 plasma units from an Italian blood donor centre, where machines are set for the collection of 700 g of plasma. Next, we built a statistical model to predict how the final volume of the unit should be set to obtain 50%, 80% or 90% probability of having at least 600 mL net plasma. RESULTS: ACD-A volume was associated with haemoglobin, total proteins and triglycerides. Donors with low haemoglobin reach an 80% probability of at least 600 mL net plasma with units of approximately 690 g, while 720 g are needed for donors with high haemoglobin levels. For total proteins and triglycerides, plasma units may vary within a range of ±20 g. DISCUSSION: Our model, based on easily measurable individual characteristics, makes it possible to customise plasmapheresis procedures by determining the blood volume to be processed for each donor. Tailored plasma donations might result in both a reduction in adverse events and an increase in the quality of collected plasma.