Convalescent Plasma Therapy: Data driven approach for finding the Best Plasma Donors

Convalescent Plasma Therapy is an investigational therapeutic method recommended as a treatment strategy for COVID- 19 as vaccines, and proper treatment methods were unavailable. The therapy involves transfusing antibody contained plasma from the COVID recovered individuals (donors) into critically affected patients. It can accelerate the recovery of the recipient. The effectiveness of antibodies is affected by the health and clinical history of donors, according to research. It implies the possibility of implementing Machine Learning Classification models for predicting the Eligible donors (who meet the threshold antibody level for donation) and Regression models to predict the antibody level value of a donor from the person's clinical history before conducting tests for the same. The proposed system can help the health authorities approach the most probably efficient donors for the therapy rather than wasting time and test kits on a random donor who may or may not be eligible. The results from various ML algorithms trained on a synthetic clinical history dataset are examined and assessed as significant to some degree. The system has to be validated against real data to arrive at reasonable conclusions. This paper demonstrates how a data-driven solution is more beneficial than the conventional methods for donor search.