ii4sm will be at the SOA in Healthcare Conference in Arlington, VA, USA, July 12-14, 2010.
Join the session 1 "Business Value of Semantic SOA; Standards-based Interoperability is NOT enough" on July 12th at 10 am:
“Business Value of Semantic SOA”
(Standards-based interoperability is not enough)
Issues of efficacy of treatments and cost-effectiveness continue to plague the healthcare industry. Costs continue to spiral, medical errors continue unabated. In the US, many people have fallen out of the system and are uninsured. Other countries struggle with national programs that have yet to realize the perceived benefits.It is believed by many that the full computerization of healthcare information in a way that enables sharing of data to maximize benefit to the patient and to the healthcare profession can provide huge gains and cost savings, e.g. cut crisis-level asthma attacks by 80%, heart attacks and kidney failure by one third 1.
However, healthcare information still remains locked away in either paper records or in technological islands divided by disparate technologies, different representations (format/syntax), various semantics and levels of semantic richness, spread among organizations, payers, providers, pharmaceutical companies, research institutes etc.
The notion of interoperability and Health Information Exchange has been almost a Holy Grail in healthcare in recent years. Billions of dollars have been poured in, and continue to be poured into finding ways of exchanging healthcare information between all participants of the healthcare and life sciences eco-system.
Right now, very large sums of money are being set aside to address this challenge by Governments throughout the world. In the USA, ARRA has provided $millions. However, if the silos are opened up and healthcare information is exchanged, how much benefit will truly arise?
The real problem is that where healthcare information is actually captured in an electronic form, it is often textual or at best in a very basic syntactic format or the same information is represented in many different ways. The work to extract it and turn it into a meaningful form to be truly analyzed, and made useful in areas such as disease management, drug efficacy analysis, signal detection and in real time clinical decision making is huge if not impossible.
Our contention is that while the huge investment into interoperability is worthwhile to a certain degree, it is simply not enough to exchange information at the level of semantic awareness that is prevalent in today’s systems. Information must be captured and managed in a truly semantically rich and robust fashion in the first place. This calls for a radical, but incremental change in the way that healthcare applications are designed and constructed. This cannot happen overnight, and should be carried out in conjunction with interoperability efforts, which must allow interaction at the highest level of semantic awareness that the participating systems can manage at any given point in time.
Technology and syntactic interoperability challenges can be solved to a large extent using integration engines and horizontal as well as vertical integration standards. Messaging standards have been defined with varying degrees of success, HL7 V2, HL7 V3, OpenEHR, DICOM, NCPDP, etc. Common domain models, such as HL7 RIM and BRIDG, and clinical terminologies, such as SNOMED, can improve the semantic awareness to the next level.
However, the picture is still incomplete. Standards are typically consensus based, co-operative efforts resulting in the least common denominator, and often leave a great deal of room for interpretation and ambiguity. To take one simple example, there are usually multiple ways to represent the same concept, e.g. “severe headache”, “pain” in location “head” with degree “severe”.
The aim is to achieve deterministic “computable semantics”. Information must be captured in a truly semantically rich and robust fashion. The use of semantically robust systems based on well defined information models, standard complex data types and rigorously defined ontologies / terminologies, coupled with reasoning and inferencing can overcome most of these challenges, using “semantic disambiguation” and where possible, “semantic enrichment”.
The presentation will discuss the elements that are needed to achieve computable semantics and the approach being taken by ii4sm towards realizing the solution of computable semantics for healthcare information.
And meet with
Alan Honey, Chief Architect, as well as Robert Lario, Chief Technology Officer, ii4sm, Basel/Switzerland
Click here for the complete program of the conference.
1 Healthcare Reform Now (G. Halvorson) ISBN 978-0-7879-9752-6