Blog

NHS to study AI-driven blood matching algorithm

The NHS has initiated a study examining the use of an advanced algorithm to select genomically matched blood to improve safety and ensure the most efficient use of rare blood products.

In contrast to traditional blood matching tests for major blood groups (like ABO and Rh), the bloodMatcher algorithm, which was developed by NHS Blood and Transplant (NHSBT) and University College London (UCL), incorporates data from additional minor blood groups to find a more precise match between donor and patient. The system analyses extensive DNA data to automatically select the most compatible units from blood banks, thereby reducing the likelihood that a patient’s immune system will develop antibodies against foreign antigens in the donated blood, which is a major concern for patients with inherited anaemias, thalassaemia, or sickle cell disorder. This can lead to transfusion reactions and make subsequent transfusions more difficult and higher risk.

The new study will evaluate bloodMatcher against blood selected manually using the current policy, where blood is normally only matched for people with sickle cell against four core blood groups.

The feasibility study is being led by NHSBT, working with University College London Hospitals NHS Foundation Trust (UCLH) and the National Institute of Health and Care Research (NIHR) UCLH Biomedical Research Centre, with additional funding from the NIHR AI programme.

Following informed consent, researchers will use the algorithm to guide the selection of blood products for around 40 adult participants who receive regular transfusions for sickle cell disorder at UCLH. The study will pave the way for a larger multi-centre trial. 

The research will be lefd by Dr Sara Trompeter, Consultant Haematologist at NHSBT and UCLH. She said: “Currently, given the numbers of units available and orders to fulfil, it is not practical to consider many blood types – the blood types are too complex, and the number of units too numerous for humans to compute the best unit to send. 

Reducing mismatch

“This new technology hopes to reduce the potential mismatch between donor and recipient minor blood groups. Mismatches can cause the formation of a blood group antibody which makes it hard to find blood the patient can safely receive or even severe transfusion reactions.

“The population cohort that most struggles with the complications of alloimmunisation is the study population of interest in this study – adults with sickle cell disorder. Around 17% of adults with sickle cell are alloimmunised.

“The new bloodMatcher algorithm will – combined with the new genomic blood data we have available– be able to select units in a faster, far more advanced way. The goal is to improve donor to patient antigen matching and thus reduce the risk of alloimmunisation and the risk of severe transfusion reactions for patients needing blood. By improving matching, we will de facto also reduce ‘waste’, caused by giving valuable rare units to people who don’t need them.”

John James OBE, Chief Executive of the Sickle Cell Society said: “Blood transfusions are an important part of treatment for many people living with sickle cell disorder. But blood needs to be closely matched, otherwise transfusions can become more difficult over time. We welcome this study, which aims to provide blood that is even more precisely matched to the person receiving it.

“For people living with sickle cell, developing complications from transfusions can make treatment more complex and more stressful over time. Research into improving the precision of blood matching could lead to a big improvement in treatment experiences, patient safety and long-term care.”

About the author

Asonblog

Add Comment

Click here to post a comment