Personalised Simulation Technologies for Optimising Treatment in the Intensive Care Unit

Summary

Participation Requirements

Description

In the United Kingdom, approximately 142,000 people are admitted to ICU each year. A large proportion, 10 - 20%, of these patients have a life-threatening respiratory illness called Acute Respiratory Distress Syndrome (ARDS). These patients need specialist help with their breathing, from a machine c...

In the United Kingdom, approximately 142,000 people are admitted to ICU each year. A large proportion, 10 - 20%, of these patients have a life-threatening respiratory illness called Acute Respiratory Distress Syndrome (ARDS). These patients need specialist help with their breathing, from a machine called a ventilator. Only seven out of ten patients will survive this illness and even survival may bring ongoing problems, sometimes for a long time after leaving hospital. Accurate mathematical and computer models of ARDS, would allow investigation of the illness outside of the ICU and inside the virtual environment of a computer. Different treatments could be simulated on the same 'virtual' patient, or the same treatment on many different patients with varying degrees of illness. Development of these software models, requires collection of a library of data describing how patients respond to changes in their treatment. An example would be to describe how a patient's blood pressure responds to a change in the settings of their ventilator. The changes to a patient's ventilation would be made as part of the normal care provided by the doctors and nurses looking after them. Mathematical descriptions have been created before, from simpler data sets which were essentially single snapshots of a patient's condition and treatment. The investigators aim to capture sequences of snapshots over several hours, allowing them to build more accurate models. Guy's and St Thomas' NHS Foundation Trust (GSTFT) is the clinical partner of the project. Patients would be identified there by clinical researchers, who would then collect the data describing their treatment. This data would be anonymised before adding to the library of data to be shared with academic researchers. Academic members of the team at the University of Warwick and the University of Nottingham possess the engineering and mathematical expertise needed to develop the complex software models. They also provide the facility of a high performance computing cluster necessary for the difficult process of fitting models to the data. Once the software models have been built and used to examine the how treatment might be improved, the findings would be shared with clinical staff around the world, through the publication of articles in medical journals. It is possible that the insights gained by the modelling process might inform, change and improve how clinical staff use ventilators to support patients with ARDS.

Tracking Information