Using artificial intelligence to inform treatment decisions: an app to help doctors care for patients with lupus

Study code

Lead researcher
Dr Max Noble

Study type
Data only

Institution or company

Researcher type

Speciality area
Dermatology, Musculoskeletal Disorders


Systemic lupus erythematosis (SLE) is an autoimmune condition that affects approximately 1 in 1000 people in the UK, according to the charity LUPUS UK. SLE impacts people in many different ways, meaning that the symptoms one person experiences may be different to those of another. The same person might also experience many changes in their condition over time. This makes SLE a very complex disease to manage; a challenge made yet more difficult by few effective treatment options.

To improve care, researchers have been collecting and analysing patient data to detect patterns, group patients according to similar characteristics and predict what might happen next. Unfortunately, due to the variable nature of SLE, this has been very difficult, resulting in a lack of clarity significantly impacting on diagnosis, management, quality of life and ultimately, patient health outcomes.

Although there are a number of new treatments in the pipeline, we believe achieving improved health outcomes will require what is known as a ‘precision medicine’ approach that takes into account the complexity of the disease and every patient’s unique experience of it. As such, we are developing a decision support application for health care practitioners to use in the clinic. This application, or app, will use machine learning techniques to bring together information that will support clinicians with a decision framework with the aim of providing predictive indicators as to whether or not a given treatment may be effective, and ultimately make the best choice for the patient. This information will include demographic, social, clinical and genomic patient factors and by its nature is highly variable. The aim is to increase the predictive power of the decision algorithms overtime, as with any clinical decision support system.

Potential patient benefit:

This activity will contribute to the development of an application that will facilitate clinical decision-making based on the integration of multiple factors known to influence therapeutic response in SLE. Patients will benefit from a more informed and personalised treatment pathway, brought about by machine learning approaches that will support the complex nature of clinical management, aiming to give 'the right treatment to the right patient at the right time', particularly as new modalities are added to the treatment armoury over the coming 5 years.