Melbourne-based medical research organisation BioGrid Australia has developed an e-health application which promises to break down information siloes between institutions by offering it as an integrated service.
BioGrid aims to provide an innovative medical research platform that facilitates “privacy protected research” across hospitals and medical research organisations.
BioGrid director Dr Marienne Hibbert said developing the portal provided an interesting example of consumer involvement in an IT project.
“The issue around consumer health is if your information is locked up in a clinic or hospital,” Hibbert said.
“If you have a personal record with important information it needs to be presented in a way people can understand.”
Hibbert said BioGrid is a distributed system aimed at collecting clinical data at the source, for example at a hospital, collecting it and making it accessible to approved researchers.
The application was developed over five months with IBM’s WebSphere portal environment.
The first practical use of the application is the CART Wheel portal which is tracking the symptoms and treatment of rare tumors.
“CART wheel was driven by a woman whose son was diagnosed with a rare form of cancer,” Hibbert said. “The patient enters data and consents to us contacting them to verify their pathology. The whole point of the program is the rare cancers make up 20 to 30 per cent of all cancers, but only 1 per cent of funds go to researching them so it is difficult to get a critical mass to mount a study.”
BioGrid’s flagship project, the Australian Cancer Grid (ACG), received $11 million from the Victorian government. Federal government funding is also provided for the not-for-profit.
“We want to increase the numbers on clinical trials and attract engagement for people with rare tumors,” Hibbert said.
“It's like a match making exercise between researchers who want to improve the health and patients who would like to participate in potential treatments.”
Hibbert said the application is generic and can be applied to treatment of diabetes, Asthma, or “just about every chronic illness”.
“Mild Asthma becomes more serious as people get older and even if it is occasional to record that information so there may be evidence of a health condition,” she said.
“One of the biggest things in health is there is always the assumption that the data exists. In hospitals the main data exists as the ‘intern level’ data, but what often doesn't exist is the sub-speciality or outcome data. And for Diabetes, Asthma or Cancer often it doesn't so we have to develop systems to collect that data. We bring together aspects of disease and clinical outcomes data.”
BioGrid is now looking to set up the application as a service and Hibbert said adoption by state or national health departments would be the best outcome.
“Another good outcome would be the federation of hospital data,” she said.
“Our privacy-protected federated model is very interesting to any industry with those sensitivities and I have spoken to fire and police departments. This technology can connect to whatever is there so it doesn't matter what the database is, there is a wrapper.”
Interoperability key to better e-health
With BioGrid being a collaboration of hospital and university members, its applications are designed for interoperability.
Hibbert said a lot of work has been done on the privacy of cross-institutional reporting and BioGrid will work with “whatever standards are there”.
“We are talking to NEHTA and they are using the SNOMED standard, but not everything that exists in health is in that standardised form and not everything that clinicians want to collect is fitting in that standard,” she said.
“You can be a mapping and translational service, but there is a wealth of information that already exists so we need to be able to communicate with existing data.”
“At a hospital one of the issues is they will have their own clinical viewer across every single enterprise system so it is point-to-point. The clinicians might think it’s an enterprise view of the patient, but it is quite siloed. Others will create a warehouse around clinical systems.”
Hibbert said hospitals claim to have an e-health systems, but they may not have all the required sub-specialty data.
“Even with a hospital system you need a clinical view, but also the research and quality side of that,” she said. “Down the track we will get there.”
“What is really needed is a clinical viewer across hospitals and we have done a proof of concept around that. That is a huge step forward for clinicians having access to clinical data of patients. When a patient goes to another institution does the clinician have access to their records?”
BioGrid is conducting an e-heath and data exchange trial among health institutions in Victoria for the treatment rectal Cancer.
Hibbert said the same system is relevant in the treatment of newborn babies as the relevant information needs to be available to the right clinician, be they a midwife or obstetrician.
“People are not used to having their own information, but as you get older, or you have a chronic illness, things need to be managed more,” she said. “Your health is a partnership between you as a patient and your health care provider. All of that involves being able to record symptoms and we have the technology to get data from hospital and GPs into a records.”
Having a history that is shareable across GPs or hospitals so the patient is “taken seriously” is also important, Hibbert says.
“Eventually the vision is wherever health data is generated the data custodian can have access to it and people can collaborate to improve health care and outcomes,” she said.
Hospitals and GPs have legal and custodial control over a person’s health information, but Hibbert says consumers need to have “good information” about their own data.
“In the future linking genomic data with clinical data will be able to drive the information cycle,” she said.
“With our current health system reporting might be three or six months down the track so data analysis is not timely.”
BioGrid uses a number of analytics tools, including SAS and SPSS (owned by IBM) as the researchers “need variety”.
“The only thing that matters is you are always comparing apples with apples. The data must be able to be understood,” Hibbert said.