iCIMS brings native interoperability to clinical specialties
Well-known health information specialist Jon Patrick has developed what he calls a new class of technology to address the need for custom-designed information systems for clinical specialties.
Since leaving the paid service of the University of Sydney in July last year, Professor Patrick (pictured) has established two new companies, one called Health Language Laboratories (HLL) and the other Innovative Clinical Information Management Systems (iCIMS).
Professor Patrick said HLL was developed to use his experiences to create industrial language processing products, which have been used in a recent project for the Victorian Cancer Registry involving information extraction and information inferencing for radiology reports.
“Radiology reports are particularly interesting because no one around the world seems to be using them for cancer registries,” he said. “It's innovative and it's a wealth of knowledge about cancer which is not being exploited in terms of the population distribution of cancer.
“It brings a much better focus on recording a range of cancers that you never get pathology reports for, such as brain cancers and benign cancers. You also get the opportunity to get a staged diagnosis … so you are getting cancer information much earlier in the cancer journey.”
With iCIMS, Professor Patrick is attempting to create a new class of technology for clinical systems produced at a much lower cost than hospital solutions from the large vendors. He believes that clinical solutions have to be readily changeable, and “it has to be clinical team-designed and it can't be designed by the vendors,” he said.
The technology consists of a design tool that can be used by the clinical team to create a system that suits their needs and can interoperate with enterprise-level clinical information systems such as EMRs and pathology and radiology systems through HL7 messaging. The technology also enables the clinical systems to be changed in real time.
Clinical teams can use the design tool to create a system, with the support of iCIMS process analysts, that does not require any further programming between the design stage and creating the run-time system.
“There's no programming, no data table design, no SQL design, nothing,” he said. “It represents a very significant change that we call an emergent clinical information system.”
The new technology uses what Professor Patrick has coined “native interoperability” to call up data from other clinical systems, and it can be used to develop systems for many different clinical specialties, as well as patient flow purposes such as theatre and ward management and to design clinical and scientific registries.
Professor Patrick describes an emergent CIS as the notion that what a clinical team wants emerges from the design process and it may have behaviours that are not part of a pre-specified design.
“They may have emerged from the way they put things together and the whole is greater than the sum of the parts,” he said.
“The other important part is that it completely moves you out of the waterfall model of system design – you are no longer in that process at all. What you have is a continuous iteration of design and test and then commission, and that's it.”
One of the benefits is that clinicians can make changes to the system in real-time without having to ask an IT vendor to do it for them. If it is a major change then iCIMS' process analysts will help, but if it is merely a new field that needs to be added or deleted, the clinical team will have the ability to alter their system themselves.
“As part of the contract they have a copy of the design tool. That gives them a great deal of confidence even if they never use it. They have the confidence that they have control over their own system, and they can ring us if they want us to do it or they can do it themselves.”
This is important for clinical specialties, which often have their own idiosyncratic workflows and different needs from site to site. This often makes it almost impossible even for the best-of-breed vendors to design a suitable system for clinical specialties aside from ED and ICU, he said.
He emphasised that there is a significant effort required to learn how to use the design tool as it has very rich functionality, but that it does allow changes to be made in real time and therefore is much more agile than off the shelf systems.
The iCIMS system is suitable for a large range of clinical specialties, such as emergency department, gynaecological oncology, breast cancer and trauma, as well as the creation of tumour registries and theatre, surgical ward and general ward management.
“Our operations model consists of all of these clinical specialty systems that are highly changeable, highly adaptable, which is what we deliver,” he said.
iCIMS sits in the middle of the information flow process, with hospital-wide clinical services systems such pathology, radiology and EMR sitting on the top. These systems have to support users throughout the hospital and therefore the capacity for them to be specialised is low, he said.
“Then you've got the clinical departments who need their own specialised systems. There are two other groups related to the clinical departments – there are the research groups, which often run their own systems in spreadsheets or Access databases and are an unholy mess in terms of an organisation – and then you have registries.
“There are three types of registries under our model. There are the population-based registries like the Victorian Cancer Registry; there are the clinical specialty or professional society registries, which are registries of what the members of those professional societies do so they can monitor the outcomes of their work; and then there is a third group that we call scientific registries.
“Scientific registries are set up to answer specific scientific questions. They collect particular data so they can make scientific conclusions.
“The research information systems and the registries both want the clinical data, so we have a model whereby you have the clinical specialty systems and they feed data to their respective research groups and their registry groups.
“There is a critical design component of our system that makes all of that doable in a highly efficient way, called native interoperability. The notion of native interoperability is that when you build your own clinical system in our environment, you can refer to any field in any other system.”
The iCIMS system uses a graphical user interface (GUI) rather than a script, and the programming language allows users to call up information by reference.
“What it means is that when you set up a research system, you have all of your own stuff in there, and then you want stuff from a clinical system, so you set up a form for them, and the GUI allows you to say where you want to get it from and the moment it is collected there it is transported here.
“So now we have auto-population of registries, auto-population of research systems, and we have auto-population of other clinical systems.
“There are three key innovations in the iCIMS system: clinical team-led design, native interoperability and real-time change. Because the clinicians design their own solutions in an iterative way, what you have is an emergent system. It has properties and behaviours beyond the design itself.”
The technology has been used in a proof of concept trial for the Victorian Comprehensive Cancer Centre, with a gynae oncology system designed for Melbourne's Royal Women's Hospital and a breast cancer system for the Royal Melbourne Hospital.
iCIMS has also built an ovarian cancer system for David Bowtell at the Peter MacCallum Cancer Centre and a cervical cancer system for Kailash Narayan, also based at Peter Mac, that are fed data automatically using native interoperability from the gynae oncology system installed at the Women's.
Professor Patrick and his team are currently in discussions with several hospitals for wider use of the technology, including the potential to design a full CIS for an emergency department. He is also in the unpaid service of Sydney University, continuing to hold an honorary position and supervising several students doing their PhDs.
Posted in Australian eHealth