UK Clinical Research Network : Portfolio Database User Guide  
Welcome to the UK Clinical Research Network Study Portfolio : NIHR (England)  

NB: The information displayed below does not replace the protocol. The latest protocol version should always be consulted before making clinical decisions.
Modelling the Physiology of the Critically Ill

Goal-Directed Trajectory Planning of Therapeutic Pathways for Septic Shock Patients Using Hybrid Knowledge Maps

Portfolio Eligibility
Automatically eligible
UKCRN ID 12855
Research Summary
The aim of this project is to develop a new type of engineering design which is aimed at modelling what happens to patients when they are critically ill due to sepsis. Early identification of deterioration and more rapid intervention, results in better outcomes. These patients may for example currently have some acceptable physiological signs such as blood pressure, but may show evidence that they are starting to tip from being ill but stable, into a phase of deterioration. Therefore if such a “map” could be developed it would have the potential to give early warning of which patients need urgent changes in treatment. Once designed, it will be able to refine itself over a period of time, by learning from new data. In the first instance to establish the basic design, retrospective data from 100 patients will be used to produce a map of the behaviour of critically ill septic patients, initially focussing on the respiratory and cardiovascular systems. This is a self learning neurofuzzy method. This means it can assess progress more like humans do, unlike techniques based on statistical models. The new methods allow the development of a system that will be able to suggest changes to therapy. In this phase of the study no changes will be made to the treatment of any of the patients. However, it is intended that the system will eventually be able to run in real time, allowing advice to be generated which for our purposes at this stage can then be used in a simulated environment, rather than with the actual patient. The ability to ensure timely changes in treatment for these patients is very important as currently 25% to 30% of patients ill enough to require general intensive care do not survive to leave hospital. Once the system is running in a simulated real time setting, we will compare changes in treatment that are advised by the system to the advice of ten ICU Consultants and ten advanced intensive care trainees. This process will not affect real patients.
Study Type Observational
Design Type Not specified
Disease(s) All Critical care
Phase N/A
Current Status Closed - in follow-up
Closure Date 5/29/2015
Global Sample Size 120
Global Recruitment to Date
Geographical Scope Single Centre
Lead Country England
Main Inclusion Criteria
There are no risks to patients.
We will include patients who are or have been critically ill, whose data has been recorded on the Patient Data Management System since 2000.
Although our principal interest is in severely septic patients, we will need a full range of severity of illness to ensure that the system learns not just to predict more severe outcomes.

We will include intensive care staff who will be presented with virtual information from the system so that we can compare their responses to different levels and types of illness against the simulated advice of the system.
Main Exclusion Criteria
Patients with incomplete data.
Any patients who have asked for their data not to be included in research or audit.
Chief Investigator(s)
Prof Mahdi. Mahfouf
Further details, please contact
Prof Gary Mills

Sheffield Teaching Hospitals NHS Trust
Northern General Hospital
Herries Road
South Yorkshire
S5 7AU

Tel: 0114 271 2381
Funder(s) Engineering and Physical Sciences Research Council (EPSRC)
Sponsor(s) Sheffield Teaching Hospitals NHS Trust
If you experience problems using the application, please click here.