Predict Blood

A new website offering an accurate prognostic tool for patients with myeloproliferative neoplasms (MPN)

Technology

A new website offering an accurate prognostic tool for patients with myeloproliferative neoplasms (MPN), a myeloid cancer that develops from a stepwise transformation of haematopoietic stem cells. While chronic MPN can generally be managed for years, a subset of patients will go on to develop disease progression and more aggressive blood cancers such as AML. Predict Blood is the new website to help accurately identify these patients and enable clinical monitoring.

Advantages

• Scientifically validated. Original study.
• Utilised by peers –  https://onlinelibrary.wiley.com/doi/full/10.1111/bjh.16924
• UK CA Marking as an MHRA Class I medical device.
• Present in national guidelines (https://pubmed.ncbi.nlm.nih.gov/37932932/).
• Allows for a personalised prediction of outcome for patients.
• Allows exported prediction reports for clinical use.
• Enhanced user experience, communication and functionality with a new interface.
• Responsive graphics – improved usability across different platforms such as mobile
• devices.
• Space for Development – new features can be added and tested efficiently.
• Robustness – Continuous Integration (CI) testing allows for automatic testing at any given time.
• Deployment – well-documented, consistent and simple procedure to update versions.

Applications/Context

• Potential to guide monitoring of patients by healthcare providers. 
• Potential to develop features that can allow patient stratification for clinical trials.
• Potential for building further tools for other blood cancers and disorders to better identify those at risk of poorer outcomes.

Comparable Technologies

Similar algorithms available are able to stratify patients into low, intermediate and high risk groups. Predict Blood provides an individualised prediction.

Background

Predict Blood is the new website of a personalised prognosis tool for patients with a particular type of blood cancer. The original model was developed from over 2000 patients who were part of a research study looking into how variations in the genetic code of MPNs and their clinical features affect disease progression and patient survival. 

Researchers chose patients on the basis of high-quality clinical data and test results. Patients were predominantly UK-based and were followed for many years. This information allowed the original algorithm underlying the new website Predict Blood to be developed and provide a prediction of prognosis in new patients. 

The model behind Predict Blood is a multistate survival model incorporating 63 clinical and genomic variables, that estimates a patient’s probability of transition between stages (or outcomes) of disease.

Intellectual property

Predict Blood is open for use. It is based on the input of genetic information and should only be used in conjunction with a suitably qualified health professional.

The Wellcome Sanger Institute is offering non-exclusive licenses to the algorithm for further development or commercial use.

Download

Back to Technologies