Measuring the Impact of a Blood Supply Shortage Using Data Science.

Background: Transfusion medicine is the only section of the clinical laboratory that performs diagnostic testing and dispenses a drug (blood) on the basis of those results. However, not all of the testing that informs the clinical decision to prescribe a blood transfusion is performed in the blood bank. To form a holistic assessment of blood bank responsiveness to clinical needs, it is important to be able to merge blood bank data with datapoints from the hematology laboratory and the electronic medical record.

Methods: We built an interactive visualization of the time from hemoglobin result availability to initiation of red blood cell (RBC) transfusion and monitored the result over a 2-year period that coincided with several severe blood shortages. The visualization runs entirely on free software and was designed to be feasibly deployed on a variety of hospital information technology platforms without the need for significant data science expertise.

Results: Patient factors, such as hemoglobin concentration, blood type, and presence of minor blood group antibodies influenced the time to initiation of transfusion. Time to transfusion initiation did not appear to be significantly affected by periods of blood shortage.

Conclusion: Overall, we demonstrate a proof of concept that complex, but clinically important, blood bank quality metrics can be generated with the support of a free, user-friendly system that aggregates data from multiple sources.