The incidence of head injury is estimated at 1500 per 100,000 population per annum. On a European dimension this translates into over a million head injuries per year. Head injury is a leading cause of death in young males and many survivors have serious and long-term morbidity. The loss of employment to the victim and the stress and increased burden of care to family members have significant social and economic effects upon Europe.
Despite many promising multi-centre trials of new drugs and management methods, very few have shown conclusive evidence for improving the survival and outcome of brain injured patients. There are potentially many reasons why we have not been successful in treating this group of patients, but many believe that the large variation across centres in how patients are managed during their acute intensive care and the continued reliance upon low-resolution paper based methods (which is likely to under-estimate the burden of secondary insults) has only hindered progress in this medical domain.
The BrainIT group works towards developing and testing more “Information Technology” based tools for collection and analysis of better standardised and higher resolution data from head injured patients. This concept has been developed and now the BrainIT group (Coordinated from Glasgow by Dr Ian Piper) has won three major EC project grants which has fostered the creation of the BrainIT network.
The group have created a standardised IT based core dataset collected from 22 neurointensive care centres from 11 countries across Europe. They have amassed and share a uniquely detailed database of over 250 patient’s data from which analyses have lead to a number of peer reviewed academic publications. The most recent EC funded project (AVERT-IT) has developed a novel “Neural Network” technology for prediction of arterial hypotension (low blood pressure) a common adverse event that occur to patients while being managed in intensive care. This new AVERT-IT technology has been fully tested in a multi-centre controlled trial. To provide a suitable real-time data collection environment for this trial - the AVERT-IT project has enabled us to develop, implement and support “ICUnet”. This unique state of the art clinical research and audit infrastructure is described in more detail below.
We believe that the BrainIT groups approach to developing data collection and analysis standards, sharing data for hypothesis testing and developing new treatment and analysis strategies will eventually lead to improvements in the acute care of patients with brain injury.
“ICUnet” is a unique Grid based security focused infrastructure setup across 6 BrainIT centres (Glasgow, Barcelona, Heidelberg, Monza, Uppsala and Vilnius) which provides automatic collection of intensive care monitoring data, laboratory data and other routinely collected clinical data for either patient management or medical research. The ICUnet system was developed, implemented and supported by a collaboration between Anthony Stell within (Professor Richard Sinnot's research group) and Rob Donald formerly of C3 Global Ltd now working as an independant consultant (Stats Research UK). A local ICUnet server sits behind hospital firewall and supports interfaces with bedside monitoring and other clinical data sources. High resolution data is sampled on-line, parsed into the BrainIT common data elements definition and stored in a local SQL server database. Secure Grid services running on-line strip off patient identifiers, encrypt data and once ever 60 minutes BrainIT data is pushed outside of hospital firewalls to a secure data service housed at the National eScience Centre at Glasgow University. This unique system supports quasi real-time remote monitoring of hospital data collected for research or clinical audit purposes. Virtual private network connections into each of the hospital servers provide software maintenance and fault tracking. BrainIT is one of the first international academic collaborative networks to develop, implement and use such a powerful research and clinical network. ICUnet, although currently implemented with fixed hardware servers can also be deployed with other architectures. We are currently assessing a cloud based solution for providing the same services as ICUnet.
The figure above shows the ICUnet system architecture in one of six BrainIT centres. Staff ICU nurse log patient's onto local information systems. Research Nurse log patient's onto the study (in the case shown it is the AVERT-IT study) via an ward web app (WWA). This triggers “Clinical Client” software to parse clinical, laboratory and monitoring data from local hospital based systems and push data onto a local SQL database. Non-hospital Research data is entered via the WWA program. For the AVERT-IT study, the Hypopredict BANN Engine software calculates probabilities of hypotensive events from data acquired from the local SQL database. Once per hour “Data Push” software anonymises local data and pushes clinical, monitoring, Hypopredict and status log files out of hospital firewalls and up to a central SQL database sitting at the National eScience (Nesc) centre in Glasgow University. A trial monitor can view and run queries on study data arriving from all six BrainIT centres in quasi-real time.
BrainIT has a "Steering Group" who meet regularly to discuss group projects and database analyses.
|Ian Piper.||Glasgow, Scotland||Group Coordinator|
|Iain Chambers.||South Tees, UK|
|Giuseppe Citerio.||Monza, Italy|
|Bart Depreitere.||Leuven, Belgium|
|Rob Donald.||Dingwall, Scotland|
|Per Enblad.||Uppsala, Sweden|
|Barbara Gregson.||Newcastle, UK||Group Statistician|
|Tim Howells.||Uppsala, Sweden|
|Laura Moss.||Glasgow, UK|
|Jan Oliver Neumann.||Heidelberg, Germany|
|Pelle Nillson.||Uppsala, Sweden|
|Arminas Ragauskas.||Kaunas, Lithuania|
|Juan Sahuquillo.||Barcelona, Spain|
BrainIT has an "Advisory Group" who assist the Group with Scientific and Technical Developments.
|Charlie Contant.||Boston, USA||Statistical Consultancy|
|Martin Shaw.||Glasgow, UK||Programming, Web Design, Mathmatical Modelling|