Mendelscan: A digital rare disease case finding approach in primary care
Introduction to project
Working with primary care, this project aims to utilise artificial intelligence (MendelScan) to help healthcare professionals more quickly find rare disease patients and diagnose them earlier.
What is the condition?
There are a number of rare diseases that may be flagged by MendelScan software.
Rare and hard to diagnose diseases affect as many as 1 in 17 people worldwide, but diagnosing rare diseases can be challenging and expensive for the health care system. It takes 5 years, on average, to diagnose a rare disease patient in the UK.
What are we doing?
We have been working with two large primary care networks (PCNs) in the Central & South region to pilot this project – one based in the Midlands and one in Oxford. A nurse and GP within the PCN were trained with foundation knowledge of genomics to deliver this care.
MendelScan software has looked at (pseudonymised) patient records at each PCN, identifying combinations of patient codes (i.e. symptoms) that may add up to likelihood of the presence of a rare disease. MendelScan flags any patients that may be in this category where the rare disease has not been diagnosed, and recommends best next steps for diagnosis.
The Midlands PCN serves a patient population of 65,000 and the Oxford PCN serves 40,000 patients. Across the two PCNs, more than 100 patients to date have been identified via MendelScan. Patients will be invited for an appointment with the trained nurse/GP on their individual report findings and any referrals that might be necessary.
What training is available?
Mendelian – A Framework to help discussions with flagged patients.pdf
Where can I find out more?
Who can I contact?
Programme Manager – Lisa Dew